Effects of incorporation of immunoglobulin G and complement component C1q on uptake and degradation of Alzheimer's disease amyloid fibrils by microglia.

Microglia are macrophage-like immune system cells found in the brain. They are associated with Alzheimer's Disease plaques, which contain fibrillar beta-amyloid (fAbeta) and other components such as complement proteins. We have shown previously that murine microglia bind and internalize fAbeta microaggregates via the type A scavenger receptor, but degradation of internalized fAbeta is significantly slower than normal degradation. In this study, we compared internalization by microglia of fAbeta microaggregates to that of anti-Abeta-antibody-coated fAbeta (IgG-fAbeta) microaggregates and found that the uptake of the latter is increased by about 1.5-fold versus unmodified fAbeta. The endocytic trafficking of IgG-fAbeta is similar to that of fAbeta microaggregates, following an endosomal/lysosomal pathway. We also compared the internalization of fAbeta microaggregates to that of complement protein, C1q-coated fAbeta microaggregates, and found that the levels of uptake are also increased by about 1.5-fold. Rates of degradation of both types of modified fAbeta microaggregates are unchanged compared with unmodified fAbeta microaggregates. We demonstrated by blocking studies that internalization of IgG-fAbeta is mediated by Fc receptors. These data suggest that, in vivo, several different microglial receptors may play a part in internalizing fAbeta, but the involvement of other receptors may not increase the degradation of fAbeta.

Essential features of the sequence of events leading to progressive neuronal degeneration in Alzheimer's Disease (AD) 1 remain largely unknown. The disease is characterized by the presence of two types of senile plaques: classical, dense core plaques and diffuse plaques. Dense core plaques are complex lesions of the cortical neuropil containing several abnormal elements. The major component of the dense plaque is fibrillar ␤-amyloid (A␤), a fragment of 40 or 42 amino acids derived from a larger membrane-spanning glycoprotein called ␤-amyloid precursor protein (␤APP) (1). Additionally, astrocytes and activated microglia are frequently associated with dense plaques (2)(3)(4)(5).
Different forms of A␤ peptide have varying solubilities. A␤ is constitutively produced in cultured cells (6,7), and soluble A␤ (sA␤), presumably in a protein complex, is present at similar concentrations in normal and AD cerebrospinal fluid (6,8).
These observations indicate that A␤ production is a normal physiological process in vivo and in vitro. Soluble A␤ can spontaneously form into the fibrillar ␤-pleated sheet (fA␤) conformation similar to amyloid fibrils found in AD brain. The formation of fibrils in vitro depends on concentration, time, pH, temperature, and ionic strength (9 -13).
Although fA␤ forms the core of the AD plaque, several other proteins have also been found there. These include ␣ 1 -antichymotrypsin and other acute phase proteins (14 -16), complement components (17, 18), and apo-lipoprotein E (19). There is also evidence for the presence of antibodies in the brains of AD patients. One study (18) detected small amounts of IgG and light chains (kappa and lambda) in the corona of degenerating neurites around the central amyloid core of the senile plaque. Moreover, studies show the presence of antibodies that recognize microglia in the cerebrospinal fluid of AD patients (20,21). The role of these proteins in facilitating assembly of amyloid fibrils or in promoting neuron damage is unclear.
It has long been recognized that reactive microglia are closely associated with dense core plaques in AD, suggesting that these cells may be involved with the removal and/or the production of A␤ fibrils. Similarly, activated microglia are also found associated with plaques that develop in a transgenic mouse model of AD (22)(23)(24). Microglia are phagocytic cells, and they are the main inflammatory response cells of the brain (25). The origin of microglia is still controversial, although evidence suggests that their lineage is myelo-monocytic (26,27). Most cells of monocytic lineage, such as Kupffer cells, Langerhans cells, alveolar macrophages, and microglia, express surface immune molecules and play a key role in local immunological events (28). In normal adult brain, microglia are highly ramified, quiescent cells that become activated during CNS injury (29). Upon activation, microglia are capable of releasing cytotoxic agents such as proteolytic enzymes, cytokines, complement proteins, reactive oxygen intermediates and nitric oxide (30). The presence of activated microglia in plaques and the association of complement components and acute phase proteins have led to the hypothesis that plaques are a site of chronic inflammation (31)(32)(33).
Because AD involves the net accumulation of fA␤, it is important to understand both the synthesis and clearance of fA␤. We have shown that type A scavenger receptors (SRA) (34,35) on microglia bind and internalize fibrillar microaggregates of purified A␤ (36 -38). It has also been shown that these recep-tors mediate binding of microglia to deposits of fA␤ (38). Histopathological analysis of AD brain tissue showed strong expression of the scavenger receptor (SR) on reactive microglia associated with plaques (39). These data raise the possibility that SR is involved in clearance of fA␤ from the brain. However, a recent study (40) crossed human APP-expressing transgenic mice (23) with SRA-gene disrupted mice (41) and found that lack of SRA expression does not alter the number, extent, or distribution of amyloid plaques. This indicates that, in the mouse brain, clearance of fA␤ via SRA is not a rate-determining step in amyloid accumulation.
The degradation of microaggregates of fA␤ via SRs was followed for up to 3 days after internalization, both biochemically and immunocytochemically, and it was found that microglia showed intracellular accumulation of fA␤ and slow, incomplete degradation of fA␤ (42). In further studies of fA␤ degradation by microglia, we measured its degradation over a longer chase period: over the course of 12 days (43). After a short internalization pulse, there was slow, partial degradation for 3 days, but after that time, degradation stopped. Interestingly, undegraded fA␤ was released from microglia throughout the 12 days, even though the degradative function of late endosomes/ lysosomes was intact.
Although there is evidence to suggest that microglia may play a harmful role in AD, these cells could also be protective if they digested amyloid fibrils. It has been suggested that clearance of amyloid deposits following immunization against A␤ with A␤ 42 in a mouse model for AD may occur via microglial cells (44). Schenk et al. immunized transgenic mice, which overexpress mutant human APP (23,45), with A␤ 42 . The immunizations were carried out either before the onset of AD-type neuropathologies (6 weeks of age) or at an older age (11 months) when fA␤ deposits and several neuropathologies are well established. Immunization prevented the development of plaques in the younger animals while markedly reducing the extent and progression of the plaques in the older animals. One can infer from these data that fA␤ deposits can be degraded as a consequence of immunization. Almost all of the mice developed and maintained serum antibody titers against A␤ 42 . Hippocampal sections contained unusual punctate structures that were reactive with anti-A␤ antibodies, several of which appeared to be A␤-containing cells. Similar cells in adjacent sections stained positively for major histocompatibility class II and phenotypically resembled activated microglia. Together, these data suggest, contrary to our in vitro results, that microglia are able to degrade fA␤ efficiently.
In light of these new data, we investigated whether primary microglial cultures can take up modified fA␤, and if so, whether there is an increase in degradation of this fA␤ by microglia. We prepared fA␤ microaggregates and bound anti-A␤ antibody, 4G8, to microaggregates to produce 4G8-coated A␤ microaggregates (IgG-fA␤). We also prepared C1q-coated A␤ microaggregates (C1q-fA␤). In this paper, we address a number of issues: 1) Can IgG-fA␤ or C1q-fA␤ microaggregates be endocytosed by microglia? 2) Does modification affect the rate of degradation of fA␤ microaggregates? 3) Is the trafficking of IgG-fA␤ microaggregates different from that of fA␤ microaggregates? and 4) Which receptor(s) mediate(s) uptake of IgG-fA␤? We show that the modifications of fA␤ affect receptor utilization, but the rates of degradation of both IgG-fA␤ and C1q-fA␤ are similar to that of unmodified fA␤ microaggregates.

MATERIALS AND METHODS
Isolation of Microglia-We prepared primary cultures of mixed glia from newborn mice and then isolated the weakly adherent microglia from cell monolayers according to previously described methods (37,42). To prepare primary cultures of mixed glia, we obtained neocortical tissues of newborn mice, removed the meninges, minced and incubated the tissues in 2.5% trypsin (Worthington Biochemical Corp., Freehold, NJ) and 0.01% deoxyribonuclease I (DNase I) (Worthington) in phosphate-buffered saline for 5 min at 37°C. The tissue was triturated with fire-polished pipettes in 0.1 mg/ml DNase I in Dulbecco's modified Eagles's medium (DMEM) (Life Technologies, Inc) with 10% fetal bovine serum (FBS) and penicillin/streptomycin. The supernatant was passed through a 145-m mesh, recentrifuged for 5 min, resuspended, passed through a 33-m mesh, and centrifuged one more time. Cells were resuspended in growth medium and plated in 75-cm 2 flasks at a density of about 1.5-2.0 ϫ 10 7 cells per flask. The mixed glial cultures were grown in bicarbonate-buffered DMEM supplemented with 10% FBS, 100 units/ml penicillin/streptomycin at 37°C in a 5% CO 2 humidified air atmosphere. The mixed glial cultures were grown for 1 week before microglia were collected from the flasks. The cultures were then shaken to release microglia every 3-4 days for 2-3 weeks. Microglia were harvested by orbital shaking at 150 rpm for 30 -60 min in DMEM. Cells were plated onto 35-mm tissue culture dishes, whose bottoms had been replaced with poly-D-lysine-coated No. 1 glass coverslips. We used the uptake of DiI-Ac-LDL (Molecular Probes) by microglia to assess the purity of our microglial cultures, because microglia, unlike astroglia or oligodendroglia, express SRA and take up Ac-LDL. Over 95% of the cells harvested after shaking took up DiI-Ac-LDL.
Preparation of A␤ Microaggregates-␤-amyloid-(1-42) peptide was purchased from Bachem (Torrance, CA). The lyophilized powder was diluted in sterile water to 5 mg/ml and kept at Ϫ80°C. Fluorescent ␤-amyloid was prepared by derivatizing with Cy3, an orange fluorescing carbocyanine dye (Biological Detection Systems Inc., Pittsburgh, PA) according to the manufacturer's instructions. A␤ 42 was dissolved at 1 mg/ml in 0.1 M sodium carbonate/sodium bicarbonate buffer, pH 9.3, then added to the Cy3 dye. The labeled protein was separated from excess, unconjugated dye by dialysis. This stock of Cy3-sA␤ at pH 9.3 was kept at 4°C. For all our studies on the uptake of Cy3-labeled fA␤ and unlabeled fA␤, the peptide was aggregated into small fibrillar structures (microaggregates) before being added to microglial cultures. sA␤ was initially diluted in water while vortexing, mixed well, then further diluted in labeling medium (serum-free DMEM medium, buffered with 20 mM HEPES to pH 7.4, and containing 100 U/ml penicillin/ streptomycin and 1 mg/ml bovine serum albumin (Sigma)). A␤ peptides were allowed to microaggregate for 2 h at 37°C before being added to microglial cultures.
Fluorescent Labeling of Cells and Blocking Studies-All labeling was done at 37°C. Cells were first rinsed twice with dye-free labeling medium. Labeled-fA␤ or IgG-fA␤ in labeling medium was then added to the medium for 10 min at 37°C. This was followed by rinsing and fixation with 3% paraformaldehyde freshly diluted in phosphate-buffered saline.

SDS-Polyacrylamide Gel Electrophoresis and Western
Blotting-Microaggregates were centrifuged at 100,000ϫ g, for 2 h, at 4°C. They were washed, taken up in loading buffer (containing 2% SDS and 5% ␤ 2 -mercaptoethanol), and boiled for 6 min before loading on 10% SDSpolyacrylamide gel electrophoresis (PAGE) (46). Proteins were transferred onto nitrocellulose filters (Bio-Rad), which were then blocked (5% milk powder) for 1 h. To detect 4G8, the membrane was incubated with rabbit anti-mouse secondary antibody for 1 h, washed, incubated with horseradish peroxidase-linked rabbit antibodies to mouse IgG for 2 h, washed again, and developed using the ECL system (Amersham, UK). To detect C1q, the membrane was incubated for 2 h with antibody RMC7H8, which recognizes C1q (Cedarlane), washed, incubated with horseradish peroxidase-linked donkey antibodies to rat IgG for 2 h, washed again, and developed using the ECL system. All washes were done 5 times for 10 min each.
Quantitative Fluorescence Microscopy-Fluorescence microscopy and digital image collection were performed using a Leica fluorescence microscope equipped with a Photometrics (Tucson, AZ) cooled chargecoupled device camera as described previously (47). Fluorescence quantification was carried out on this widefield microscope with a 25ϫ magnification objective to obtain a large number of cells per field. Images for qualitative purposes were obtained at a higher magnification (63ϫ, numerical aperture 1.4 objective) as described previously (48). Fields used for quantification were selected at random throughout the dish and focused using phase-contrast optics before viewing the fluorescence. Digital fluorescence images were obtained, and background signal was removed from the images as described previously (49). Objects of the size of cells and above the threshold intensity were identified using Metamorph software (Universal Imaging, West Chester, PA). Total fluorescence intensity was taken as the sum of all intensity in these objects, which was then divided by the number of cells in the field to calculate the average fluorescent intensity per cell.
Uptake and Degradation of 125 I-Labeled Ligands-Cells were grown in 24-well plates for 1-2 days before the start of the experiment. 125 I-fA␤ was prepared using the chloramine-T method (42,50). The cells were washed twice with labeling medium and then incubated with the radiolabeled A␤ 42 peptides for 1 h at 37°C in a 5% CO 2 -humidified air atmosphere. The cells were rinsed three times with labeling medium and then incubated at 37°C for 15 min to allow the release of nonspecifically bound A␤. The cells were rinsed further, twice with serum-free medium and three times with chase medium, which was the microglia growth medium, and then incubated with 1 ml of chase medium for varying times. At the end of the chase, the medium was removed, and the radioactivity in the medium was measured. The cells were rinsed and solubilized with 1 M NaOH. The chase medium was precipitated with 10% trichloroacetic acid on ice for 1 h and centrifuged at 14,000 ϫ g for 10 min to separate the trichloroacetic acid-soluble fraction from the trichloroacetic acid-insoluble fraction. Where indicated, the excess unlabeled ligands were added to the incubation medium along with the radiolabeled peptides and were maintained in the labeling medium for the entire incubation. The radioactivity of these background dishes was subtracted from the values for the control experiments. All radioactive experiments were conducted with three wells per condition and repeated at least three times on separate days.
Confocal Microscopy-Fluorescence images were obtained with a Zeiss LSM 510 laser-scanning confocal microscope equipped with a 63ϫ, NA 1.4 plan Apochromat objective (Carl Zeiss, Inc., Jena, Germany). Red and green images were acquired sequentially using the 543-nm (exciting Cy3) and 488 nm (exciting fluorescein/Alexa488) lines from a 1-milliwatt HeNe laser and a 25-milliwatt argon laser, respectively.

Microglia Internalize fA␤ and IgG-fA␤ Microaggregates-
Fluorescently labeled Cy3-fA␤ microaggregates were modified with 4G8, an anti-␤-amyloid antibody by incubation with the antibody (0 -10 g/ml) for 2 h at 37°C. To confirm the presence of the antibody in the modified fA␤ microaggregates (IgG-fA␤), Western blotting was performed. Fig. 1A shows that 4G8 is bound to fA␤. The binding of 4G8 to fA␤ becomes saturated at a concentration of 1 g/ml. Specific binding was confirmed by using an isotype matched control antibody. This antibody does not bind to fA␤ microaggregates.
Binding of C1q to IgG is the first step in the classical pathway of the complement cascade that leads to complement activation. Because C1q is found in AD brains and is reported to bind directly to human A␤ (51), we were interested to see whether C1q modified fA␤ would be internalized by microglia. We found that murine or bovine C1q did not bind directly to fA␤; therefore, we utilized the anti-A␤ antibody, 4G8, as an indirect way to couple fA␤ and C1q. We prepared C1q-modified microaggregates of fA␤ (C1q-fA␤) by incubating IgG-fA␤ with 10% FBS. To confirm the presence of C1q in the C1q-fA␤ microaggregates, Western blotting was performed. Fig. 1B shows the presence of C1q bound to IgG-fA␤ after 2 h of incubation with FBS. The maximum amount of C1q is bound when the saturating concentration of 1 g/ml 4G8 is used.
To determine whether microglia take up the modified microaggregates, cells were pulsed with Cy3-fA␤ or Cy3-IgG-fA␤ or microaggregates for 10 min, washed, and then fixed with 3% paraformaldehyde. Images were acquired on a confocal microscope and single optical sections are shown (Fig. 2). IgG-fA␤ microaggregates are internalized by microglia (Fig. 2B), and the intensity of uptake of the modified microaggregates appears greater than that of unmodified A␤ microaggregates (compare Figs. 2A and 2B). Quantification of the uptake of IgG-fA␤ compared with fA␤ is shown in Fig. 2C. Internalization of IgG-fA␤ by microglia increases as binding of 4G8 increases until a plateau is reached at a concentration of 1 g/ml 4G8. At 5 and 10 g/ml 4G8, there is no further increase in uptake. At saturating concentration of 4G8 (1 g/ml), internalization of IgG-fA␤ is about 1.5-fold greater than that of unmodified fA␤ microaggregates (p Ͻ 0.01). There is no change in uptake of fA␤ microaggregates preincubated with isotype control IgG at a concentration of 10 g/ml (data not shown).
IgG-fA␤ Microaggregates Are Endocytosed into Endosomes and Traffic into Lysosomes-We were interested to see whether the trafficking of IgG-fA␤ and C1q-fA␤ microaggregates followed a similar path to that of fA␤. To address this issue, Cy3-fA␤ and Alexa488-IgG-fA␤ or Alexa488-C1q-fA␤ were added simultaneously to the same samples; pulsed for 10 min; and chased for 0, 1, and 24 h prior to fixation with 3% paraformaldehyde. Fig. 3 (A-H) shows a representative single confocal slice for the uptake of fA␤ and IgG-fA␤ microaggregates at each time point. There is a high degree of colocalization at each of the time points, suggesting that the trafficking of unmodified fA␤ (A, C, and E) and IgG-fA␤ (B, D, and F) microaggregates follow similar endosomal and lysosomal routes. By flickering between the images on screen, the colocalization can be estimated to be about 80%. The trafficking of fA␤ (Fig. 3G) and C1q-fA␤ (Fig.  3H) microaggregates also shows a similar degree of colocaliza- 1-5). The microaggregates were ultracentrifuged, washed, and run on SDS-PAGE. The gel was subjected to Western blotting for 4G8 (A). An isotype control is shown in lane 6, where an unrelated IgG2b antibody (10 g/ml) was substituted for 4G8. Lane 7 shows 4G8 alone. C1q-fA␤ microaggregates were ultracentrifuged, washed, and run on SDS-PAGE. The gel was subjected to Western blotting for C1q. B, C1q-fA␤, prepared with the concentrations of 4G8 indicated (lanes 1, 3-7). An isotype control (10 g/ml) is shown in lane 2.

FIG. 1. A␤ microaggregates were incubated with anti-A␤ antibody, 4G8, at the concentrations indicated (lanes
tion at all time points, and the image obtained immediately after internalization is shown. IgG-fA␤ Uptake Is Mediated by Fc Receptors-To investigate which receptors internalize IgG-fA␤ microaggregates, blocking studies were carried out. fA␤ microaggregates are bound and internalized by SRs (37). It is possible that IgG-fA␤, which contain both fA␤ and 4G8 antibody, could be bound and internalized by the SR as well. On the other hand, the presence of the antibody may sterically hinder the binding site of fA␤ to SR. In this case, only Fc receptors would mediate binding and uptake of IgG-fA␤ microaggregates. Fig. 4 shows that competing ligands for Fc receptors partially block the uptake of Cy3-IgG-fA␤ microaggregates (Fig. 4G), whereas the uptake of fA␤ microaggregates is unaffected (Fig.  4C). This indicates that IgG-fA␤ microaggregates are indeed taken up via Fc receptors. Conversely, the SR ligand, fucoidan, blocks almost all of the uptake of fA␤ microaggregates (Fig.  4B), but only a small amount of IgG-fA␤ microaggregates (Fig.  4F). Quantification of these, and other, images are shown in Fig. 4I. Fig. 4J shows that the blocking efficiency of fucoidan decreases as the concentration of 4G8 in IgG-fA␤ microaggregates increases to saturation, indicating that Fc receptors are sufficient to allow binding and uptake of the modified fA␤. Fig. 4K shows that when soluble IgG is added at a concentration of 100 g/ml, the uptake of IgG-fA␤ by microglia is reduced by about 50%. Increasing the concentration of these competing ligands above 100 g/ml has no further blocking effect on uptake of IgG-fA␤ microaggregates. When ligands for both SRs and Fc receptors are present, almost all binding and uptake of IgG-fA␤ microaggregates is abolished (Fig. 4I). We have not elucidated the receptors involved in the uptake of C1q-fA␤ yet, but it is likely that complement receptors are involved.
Rate of Degradation of IgG-fA␤ Is Similar to That of fA␤-We compared the ability of microglia to degrade internalized fA␤, IgG-fA␤, or ␣ 2 -macroglobulin (␣ 2 M) by following the presence of fluorescently labeled microaggregates over time. Fluorescently labeled proteins were added into the medium at 37°C for 10 min to allow binding and endocytosis and washed away, and the cells were then incubated in label-free medium for various times. ␣ 2 M enters cells by receptor-mediated endocytosis and is delivered to late endosomes and lysosomes, where it is degraded (50). By fluorescence microscopy, these endosomes appear as bright spots distributed through the cytoplasm. Clearance of labeled proteins within the cells was indicated by the decrease in cell-associated fluorescence. After 1 h of chase time the Cy3-␣ 2 M fluorescence was significantly reduced, and by 24 h there was almost complete loss of fluorescence (Fig. 5, G, H, and I). Experiments with IgG-fA␤ (Fig.  5, D, E, and F) indicate a slow rate of degradation similar to fA␤ (Fig. 5, A, B, and C). With uptake of either IgG-fA␤ or fA␤ by microglia, there was no appreciable degradation after 1 h of chase and a reduction of only about 10% after a 24-h chase (quantitative data not shown). Experiments with fluorescent FIG. 2. Microglia internalize IgG-fA␤ microaggregates more efficiently than unmodified fA␤ microaggregates. Microglia were incubated at 37°C for 10 min with Cy3-fA␤ microaggregates (5 g/ml) (A) or Cy3-IgG-fA␤ microaggregates (5 g/ml) (B). Cells were washed and fixed with 3% paraformaldehyde and imaged using a Zeiss LSM confocal microscope equipped with a 63ϫ objective. Single slices (0.5 m) are shown. C, quantification of uptake with increasing 4G8 in the Cy3-IgG-fA␤ microaggregates. The difference between 0 and 1 g/ml IgG is significant (p Ͻ 0.01 by Student's t test). All images for quantification were acquired using a widefield microscope with a 25ϫ objective.
C1q-fA␤ reveal a rate of degradation similar to IgG-fA␤ and fA␤ (data not shown).
We wanted to compare the degradation kinetics of fA␤, IgG-fA␤, and C1q-fA␤ for a longer period of time to see if the microglia carried out more extensive degradation of C1q-A␤ or IgG-fA␤ than unmodified fA␤. We previously found that about 50% of internalized fA␤ was released after 7 days, whereas 50% remained cell-associated. Only a fraction of the released fA␤ was degraded to trichloroacetic acid-soluble material (43), indicating that most of the fA␤ that is cleared from microglia has not been degraded. In this study, we found that the release and degradation of C1q-fA␤ and IgG-fA␤ microaggregates closely followed the pattern of degradation of fA␤ microaggregates (Fig. 6). After 7 days, only 50% of internalized C1q-fA␤ and IgG-fA␤ was released into the medium, whereas 50% remained cell-associated (Fig. 6A). Of the released C1q-fA␤ and IgG-fA␤, about 50% was degraded to trichloroacetic acid-soluble material after 7 days (Fig. 6B). To ensure that the trichloroacetic acid-soluble material was peptide-associated 125 I, a chloroform extraction was performed. Over 95% of the radioactivity remained in the aqueous layer, indicating that there was not a significant amount of free 125 I in the trichloroacetic acid-soluble fraction. We considered the possibility of microglial death during the 7-day chase period. In a previous study (43) we showed that, after fA␤ pulse (1 g/ml for 1 h) and 12 days of subsequent chase in label-free medium, microglial cells were alive. The accumulation of fA␤ fibrils did not affect the viability of the cell as observed from cell number, morphology, and by using a Live/Dead viability kit (Molecular Probes, L-3224). After 12 days of chase, more than 80 -90% of the cells were viable. In the current study, after the fA␤ pulse (with modified or unmodified microaggregates) and subsequent 7-day chase, we examined the cells microscopically, and the cell number and morphology in each group appeared similar. We are confident that all the cells in this experiment have similar viability after the pulse and 7-day chase. DISCUSSION ␤-Amyloid peptides are abundantly deposited in senile plaques and have been implicated as causative agents in the neurodegeneration observed in brains of patients with AD (1). A␤ is secreted in soluble form by many cell types, and its accumulation may involve changes in pathways that lead to production of fA␤. Alternatively, changes in the degradation of soluble or fibrillar A␤ could lead to net accumulation of amyloid without changes in the rate of amyloid production. Understanding the mechanisms behind normal clearance of A␤ may provide important information concerning the pathogenic deposition of A␤ in AD brains. At some stages in the disease, fA␤ may become modified by one or more of the components often associated with the plaques. Components found associated with senile plaques include apolipoproteins, acute phase proteins, cytokines, complement proteins, complement regulators, and antibodies. Therefore, it is important to investigate how microglia interact with modified fA␤ microaggregates.
We have shown previously that fA␤ microaggregates bind to microglia and are internalized via the SR (37) and that the degradation of fA␤ is slow relative to normal degradation (42). After 12 days, 60% of internalized fA␤ was released from microglia, but only 20% of this released fraction was degraded to trichloroacetic acid-soluble peptides (43). In this study we show that fA␤ microaggregates, modified with the anti-␤-amyloid antibody, 4G8, are also internalized by microglia. There is a saturable increase in the internalization of IgG-fA␤ as the amount of bound antibody increases. The uptake can be competed by blocking Fc receptors. Microglia are also able to inter- FIG. 4. Uptake of IgG-fA␤ microaggregates is mediated via receptors that are blocked by ligands to Fc receptors, whereas uptake of fA␤ microaggregates is not. Microglia were incubated for 10 min at 37°C with Cy3-fA␤ (5 g/ml) (A-D) or Cy3-IgG-fA␤ (5 g/ml) (E-H) and then washed and fixed with 3% paraformaldehyde. Images were acquired using a widefield microscope. The cells were preincubated for 10 min on ice in labeling medium with fucoidan (500 g/ ml) (B, F, D, H) or IgG (20 g/ml) (C, G, D, H). Blocking ligands were present throughout the experiment. Quantification of this and other experiments is shown (I). Uptake of Cy3-A␤ microaggregates is represented by black columns and that of Cy3-IgG-fA␤ microaggregates by gray columns. Microglia were pulsed for 10 min with Cy3-IgG-fA␤ microaggregates, prepared with increasing amount of 4G8 in the presence of fucoidan (500 g/ml) (J). Microglia were washed and fixed with 3% paraformaldehyde. Microglia were pulsed for 10 min with Cy3-IgG-fA␤ microaggregates, prepared with 1 g/ml 4G8 (K). Competing ligand, IgG, was present throughout the experiment at the concentrations indicated. Microglia were washed and fixed with 3% paraformaldehyde. All images were acquired using a widefield microscope. nalize fA␤ modified with C1q, and the uptake, as with IgG-fA␤, is increased almost 2-fold. It has been shown that microglia degrade fA␤ slowly (42,43), leading to the cells becoming bloated with fA␤ after long incubations with fA␤. In this study, we find that IgG-fA␤ and C1q-fA␤ microaggregates are degraded at the same rate as unmodified fA␤ microaggregates. We also find that the trafficking of IgG-or C1q-modified fA␤ is similar to that of fA␤ microaggregates alone.
A recent study (44) showed that, in a mouse model of AD, A␤ immunization inhibits the formation of amyloid plaques and associated dystrophic neurites. Plaques become greatly reduced even if the immunization is carried out after plaques have started to form, suggesting that degradation of fA␤ is taking place. Immunohistochemistry with anti-mouse antibodies showed that immunoglobulin was present in the remaining plaques of the A␤-immunized mice but not the control mice. Although the understanding of the immune response, which resulted in the reduced pathology of the transgenic mice is incomplete, A␤ immunization results in the generation of anti-A␤ antibodies and A␤-immunoreactive monocytic/microglial cells in the regions of the plaques. Schenk et al. (44) propose that anti-A␤ antibodies facilitate clearance of A␤ either before deposition or after plaque formation by triggering monocytic/ microglial cells to clear A␤ using signals mediated by Fc receptors. The data we present in this paper support the hypothesis that microglia are able to internalize anti-A␤-antibody-coated fA␤ microaggregates. However, because we have shown that modification of fA␤ with IgG or C1q does change the receptor requirement, it does not affect the ability of microglia to degrade fA␤, except perhaps for increasing the rate of clearance. This suggests that the in vivo clearance is not simply due to changes in receptor binding and activation in microglia. Perhaps it could be related to a more complex immune response involving activation of microglia. Alternatively, the in vivo clearance could be due to the involvement of more than one cell type in a synergistic activation.
Huang et al. (40) found that elimination of SRA does not affect plaque formation or neurodegeneration in the transgenic mice, which express mutant human APP. The plaques in the transgenic SRA gene-targeted mice did not have a greater load of amyloid than those in the transgenic mice with SRA, indicating that SRA does not play a rate-limiting role in clearance of fA␤ aggregates. Although there is little expression of SRB in the brain (52), the possibility that low receptor numbers of SRB on microglia are able to internalize fA␤ cannot be excluded. However, the fact that microglia are able to internalize modified fA␤ microaggregates via receptors other than the SRA fits well with the possibility that a variety of receptors may be responsible for in vivo clearance of fA␤. In our primary cultures, microglia are able to use Fc receptors for internalization of IgG-fA␤ microaggregates. If microglia are involved in the clearance of fA␤ aggregates in vivo, uptake of modified fA␤ aggregates may occur through a variety of receptors. The consequences of binding and internalizing modified A␤ aggregates by microglia could vary widely, depending on the signaling  6. Microglia degrade and release modified fA␤ with kinetics indistinguishable from those for unmodified fA␤. Microglia were incubated for 1 h with 125 I-fA␤ (f), 125 I-IgG-fA␤ (⅜), or 125 I-C1q-fA␤ (OE) microaggregates (5 g/ml), after which the cells were washed and chased for the times indicated. In parallel experiments with 125 I-fA␤, 100ϫ excess unlabeled Ac-LDL (200 g/ml) was added to determine the nonspecific uptake, for which measurements were corrected. In the case of uptake of radiolabeled 125 I-IgG-fA␤ and 125 I-C1q-fA␤ microaggregates, 100ϫ excess unlabeled microaggregates and excess unlabeled Ac-LDL (200 g/ml) were added to determine the nonspecific uptake, for which measurements were corrected. In all cases, nonspecific binding was not more than 12% of control binding. For each chase time, the medium was collected and the cells were solubilized in 1 M NaOH. The cell-associated radioactivity was measured (A). Chase medium was subjected to trichloroacetic acid precipitation to monitor degradation. The radioactivity of trichloroacetic acid-soluble (B) and -insoluble (C) fractions are shown as the percentage of total counts/min. The data presented are averages of the radioactive counts from at least three experiments with each sample in triplicate. pathway that may be triggered by the particular receptor that is bound.
We have shown in vitro that microglia can endocytose modified and unmodified fA␤ microaggregates. In vivo, this may be beneficial or harmful depending on the stage of the disease. A␤ has been shown to activate the complement pathway by binding to C1q, and this may result in an inflammatory reaction in an attempt to clear the stimulus. If A␤ continues to build up, the activation stimulus remains and the inflammation could become a chronic inflammatory site rather than a useful defense mechanism. In addition, fA␤ modified with other plaque components may bind to alternative receptors on microglia, which may activate the microglia and increase the rate of degradation of fA␤. In the case of the Fc receptor, uptake of IgG-fA␤ does not promote rapid degradation in the primary contacting cells.
In summary, we have found that microglia are able to internalize fA␤ microaggregates modified with IgG or C1q. The rates of degradation in each case over a 1-week period are the same, that is, slow and partial. Any modification that results in greater clearance or degradation of A␤ could lead to attempts to enhance the function of that particular receptor in the hope that this might diminish accumulation of amyloid plaques and consequent damage in individuals with AD. The fact that microglia do partially degrade fA␤ and modified fA␤ suggests that it is possible to increase this activity so that microglia could more efficiently protect against this accumulation. Further work will be required to study uptake of fA␤ microaggregates, modified in different ways, to identify the receptors on microglia by which they are internalized and to investigate the consequent rates of fA␤ degradation.