Scavenger receptors (SRs) are a family of receptors displaying affinity for a wide variety of ligands including modified lipoproteins. SRs play a range of physiological functions including intracellular transport, lipid transport and pathogen clearance. The role of SRs has been documented in pathologies such as atherosclerosis and Alzheimer’s disease. Although most studies on SRs have focused on macrophages, these receptors are also expressed in endothelial cells, smooth muscle cells and in cells of the CNS. Within the brain, SRs have been mostly studied in microglia, due to their functional similarity to macrophages. However, in situ images from Allen’s brain atlas suggest SRs (eg. Scarb2) are abundant in neurons. In this study, we have used two fluorophore labeled, well characterized SR ligands, maleylated-BSA (MBSA) and polyguanylic acid (polyG) to probe acute cortical slices. Our data indicate that within the cortex, neurons avidly endocytose both ligands. Thus, in cerebral cortex neurons may have higher number of functional SRs on the surface than other cell-types. The functional implications of our observation are briefly discussed.

SRs were identified in the laboratory of Goldstein and Brown due to their ability to endocytose modified low density lipoproteins, modified proteins like maleylated albumin and nucleotides (polyguanylic and polyinosinic acid)^[1][2]. At present, at least 10 different classes of SRs are known^[3]. SRs are versatile receptors and play numerous physiological and pathological roles and have also been successfully used for drug targeting specifically in macrophages^[4][5]. Most studies till date have focused on immunological^[6] and cardiovascular role of SRs. SRs also play a crucial role in lipid transport. Although SRs are known to be expressed in brain tissue, limited number of studies have focused on their function in the brain. Microglia, astrocytes and brain macrophages express SRs^[7]. It has been demonstrated that SRs are involved in uptake of amyloid β (Aβ) and thus play a role in Alzheimer’s disease^[8]. Some members of the SR family are also expressed in neurons^[3], however functional studies to characterise neuronal SRs have not been extensively performed. In this study, we probed acute mouse cortical slices with two independent fluorophore labeled SR-ligands to examine their cell-type-specific uptake.

The objective of this study was to investigate which brain cell type exhibited the most efficient endocytosis of scavenger receptor ligands in the mouse cerebral cortex.

MBSA has been widely used as a ligand for SRs. We generated Texas-red labeled MBSA and tested it on a macrophage cell line RAW264.7 which is known to express scavenger receptors. Our data indicate that MBSA-Texas red was rapidly internalized by RAW264.7 cells. Internalization of BSA was negligible during the same time (Fig. 1A, B). To ensure that the internalized MBSA was indeed transported to endo-lysosomal compartment, we did a pulse chase experiment on RAW264.7 cells. Within 90 min, MBSA-Texas-red co-localized with lysotracker indicating that indeed MBSA is transported to endo-lysosomal compartment (Fig. 1C). We next examined the uptake of MBSA-Texas red on acute cortical slices generated from adult mice. MBSA was rapidly taken up by some but not all the cortical cells, counterstained with nucleic acid stain DAPI (4',6-diamidino-2-phenylindole), indicating that not all the cells efficiently internalised MBSA (Fig. 1D). We therefore counterstained cortical slices with the neuronal marker NeuN, the astrocytic marker, GFAP (glial fibrillary acidic protein) and the microglial marker Iba1. Our data clearly demonstrate that only cells positive for the neuronal marker NeuN efficiently internalised MBSA (Fig. 1E, F, G and Suppl. Fig. A). This indicates that the cells which robustly endocytosed MBSA were neurons. To confirm that MBSA enters endocytic vesicles within neurons, we acquired higher magnification images, our data reveal that even in diffraction limited confocal microscopy, clear punctate structures can be observed in somatodendritic region which likely represents endocytic vesicles (Suppl. Fig. B). In order to corroborate these observations, we repeated our experiments with another well characterized SR ligand, polyG. We examined the uptake of polyG labeled with FAM on acute cortical slices. Our data show that NeuN positive cells rapidly internalize polyG but not polyA (polyadenylic acid) indicating that cortical neurons express a large number of functional SRs on the surface (Suppl. Fig. C, D). Our imaging results indicate most cells in the cortex that internalized detectable quantities of SR ligands are NeuN positive. Thus, contrary to expectations, neurons rather than microglia or astrocytes endocytose SR ligands in the cortex. Till date neuronal endocytosis has been mostly examined in context of synapses. Our data highlights the presence of highly active receptor-mediated endocytosis in the somatodendritic region. SRs type A, B, D, F, G can bind to MBSA and SRs type A, B, G, H, I can bind to polyG^[9]. Therefore, the type of SRs in neuron should belong to one of these classes. Our study does not verify if the same neurons are capable of internalizing both ligands simultaneously. Future investigations on how many different kinds of SRs are expressed by the same neuron will shed more light on the function of SRs on neurons. Other lipoprotein receptors like ApoER2 and VLDLR are also present on neurons and play valuable function^[10]. Large quantities of lipoproteins are synthesized by glial cells^[11]. In periphery SRs play a critical role in lipid transport^[12], it is therefore likely that SRs may also play physiological roles in lipid homeostasis of the brain. SRs are also known to bind and internalize Aβ, which is neurotoxic^[13]. It may be possible that SRs present on neurons contribute to this uptake and neurotoxicity. Finally, our data indicate that, like in macrophages, SR ligands may be useful for targeting drugs and bioactive molecules to neurons.

Our study indicates that, cortical neurons express large amounts of functional scavenger receptors based on the internalization studies. Surprisingly, we detected very few non-neuronal cells within the cortex that exhibited similar levels of internalization.

One big limitation of this study is that it does not specify which types of SRs are active in neuronal cortex. Furthermore, levels of SR expression may vary with age as well as pathology. It is likely that during inflammatory and degenerative states cortical microglia may express more scavenger receptors. In future, immunohistochemistry and internalization studies in pathological conditions will be required. In this study, we used two artificial ligands of SRs, thus our study does not investigate the endogenous ligands for neuronal SRs.

In future, it will be critical to examine the endocytic process using competitive SR antagonism dependent inhibition as well as by inhibiting clathrin mediated endocytosis. Furthermore, the specific type of SR has to be identified by performing immunohistochemistry, transcript profiling and gene down-regulation experiments. It will be also crucial to identify endogenous ligands of SRs in brain. SRs may internalize modified lipoproteins and thus function in brain lipid homeostasis. Furthermore SRs are known to bind Aβ and may play a role in the Aβ mediated toxicity. Careful affinity chromatography experiments will be crucial to provide further answers.

Cell culture and internalization assay

Raw 264.7 cells were kind gift from Prof. Liwu Li, they were maintained in DMEM containing 10% FBS at 37°C and 5% CO2. For internalization assay, RAW 264.7 cells transduced with lentivirus expressing GFP were plated on poly-lysine coated glass bottom tissue culture plate. 24 h later, cells were exposed to 2 μM MBSA-Texas red for 4 min at 37°C. Cells were imaged using an inverted Zeiss laser scanning microscope (LSM 880) with a 63X oil immersion lens. For pulse-chase experiments Raw 264.7 cells were plated on polylysine coated glass bottom tissue culture plate. 24 h later cells were pulsed with 2 μM MBSA-Texas red for 4 min washed and fresh media added. Cells were incubated at 37°C for 65 minutes, lysotracker was added to the cells and they were returned to the incubator for additional 25 min. Live cells were imaged using 63X oil immersion lens in an inverted Zeiss laser scanning microscope (LSM 880).

Texas-red maleylted BSA generation

BSA (Sigma) was incubated with NHS-Texas red (ThermoFisher) at pH 8.5 in a 1:3 ratio for half an hour at 4°C. Conjugation was stopped using Tris and the protein was dialyzed against PBS (phosphate buffered saline) exhaustively. The ratio of Texas red to BSA was found to be ~1:2.2. BSA-Texas red was then maleylated using maleic anhydride following a protocol described earlier^[5]; briefly maleic anhydride was added to BSA-Texas red with constant stirring, the pH was maintained above 8.5 using 5 N NaOH. This was followed by dialysis against PBS overnight. 10 mer polyG and polyA conjugated with FAM were acquired from IDT (Integrated DNA Technology).

Acute cortical slice experiment

All animal procedures were performed in accordance with the guidelines for the animal care of laboratory animals issued by Virginia Tech. The mouse strain used was C57Bl/6. 200 μm thick coronal brain slices from P35 mice (35 days old) were cut with a vibratome, the method was adapted from Wu and Hablitz, 2005^[14]. Slices were then incubated in a 24 well plate containing 5 μM MBSA-Texas Red (5 μM BSA-Texas Red as control) or 10 μM polyG (10 μM polyA as control) in 1 ml artificial cerebrospinal fluid (ACSF, Cold Spring Harbor Protocols) at 4°C with continuing oxygen supply for 30 min to allow binding. Afterward, oxygen supply was cut off and the plate was incubated at 37°C for 8 min to allow internalization. The brain slices were washed with phosphate buffered saline (PBS) 3 times and fixed with 4% paraformaldehyde (PFA). Neuronal marker NeuN antibody (Novus Biologicals, dilution ratio 1:150), astrocytic marker, GFAP antibody (Neuromab, 1;200) and microglial marker Iba1 antibody (Wako, 1:250) were used for immunostaining as needed. The slices were permeabilized overnight with 0.2% TritonX-100 in PBS, slices were blocked for 1 h in blocking buffer (0.2% TritonX-100 in PBS with 5% horse serum). Slices were incubated with NeuN antibody diluted in blocking buffer for 1 h followed by incubation for 30 min with Alexa 488 or Alexa 546 anti-Rabbit (Abcam) secondary antibody (1:500). Slices were then washed mounted on slides using Vectashield® with DAPI.

KM is supported by a grant from National Eye Institute, R01EY024712-03.

We thank Dr. Jianping Wu (Virginia Tech Carilion Research Institute) for helping with making acute brain slices. We also thank Prof. Liwu Li (Biological Sciences, Virginia Tech) for providing us with RAW264.7 cells and Dr. Michael Fox for providing us with antibody for Iba1.

The authors declare no conflicts of interest.

All animal experiments were performed according to the regulations implemented in Virginia Tech Institutional Animal Care and Use Committee.

No fraudulence is committed in performing these experiments or during processing of the data. We understand that in the case of fraudulence, the study can be retracted by ScienceMatters.