A large portion of the astrocyte proteome is dedicated to perivascular endfeet,including critical components of the electron transport chain

Abstract

The perivascular astrocyte endfoot is a specialized and diffusion-limited subcellular compartment that fully ensheathes the cerebral vasculature. Despite their ubiquitous presence, a detailed understanding of endfoot physiology remains elusive, in part due to a limited understanding of the proteins that distinguish the endfoot from the greater astrocyte body. Here, we developed a technique to isolate astrocyte endfeet from brain tissue, which was used to study the endfoot proteome in comparison to the astrocyte somata. In our approach, brain microvessels, which retain their endfoot processes, were isolated from mouse brain and dissociated, whereupon endfeet were recovered using an antibody-based column astrocyte isolation kit. Our findings expand the known set of proteins enriched at the endfoot from 10 to 516, which comprised more than 1/5th of the entire detected astrocyte proteome. Numerous critical electron transport chain proteins were expressed only at the endfeet, while enzymes involved in glycogen storage were distributed to the somata, indicating subcellular metabolic compartmentalization. The endfoot proteome also included numerous proteins that, while known to have important contributions to blood-brain barrier function, were not previously known to localize to the endfoot. Our findings highlight the importance of the endfoot and suggest new routes of investigation into endfoot function.

Keywords

Astrocyte endfeet metabolic compartmentalization proteomics somata

Introduction

Astrocytes comprise ∼30% of the neuroparenchymal volume and are numerically equal to neurons.^1,2 Astrocytes mediate numerous roles that are distributed among their many specialized processes. Through these processes, each astrocyte contacts ∼4 neuronal somata, ∼10⁵ synapses, and 1–2 blood vessels.^3,4

Perivascular astrocyte endfeet are highly specialized, diffusion-limited cellular processes that fully ensheathe the vertebrate cerebral vasculature.^5–9 Despite their ubiquitous presence, a detailed understanding of endfoot physiology remains elusive. Endfeet have been linked to numerous functions, including nutrient uptake,^10,11 modulation of vascular tone,^12–15 brain water homeostasis,¹⁶ blood-brain barrier (BBB) differentiation,¹⁷ and glymphatic flux.¹⁸ However, their precise contribution to these functions, and even their necessity to a healthy brain, remains incompletely understood.

Progress in understanding endfoot physiology has been stymied by a limited understanding of the proteins that distinguish the endfoot from the greater astrocyte body. To date there are only 10 astrocytic proteins known to be enriched at the endfoot. These endfoot proteins include aquaporin-4,¹⁹ β-dystroglycan,²⁰ α1-syntrophin,²¹ dystrophin,²² α-dystrobrevin,²² utrophin,²³ limitrin,²⁴ laminin,²⁵ glucose transporter (GLUT) 1,²⁶ and connexin-30.²⁷ The technical challenges involved in isolating specific subcellular processes have posed an obstacle to studying the endfoot as a distinct entity.

Here, to overcome this obstacle, we developed a novel technique to isolate astrocyte endfeet from brain tissue. We then studied the proteome of astrocyte endfeet in comparison to the astrocyte somata. Our dataset expands the known endfoot-specific proteome from the previously reported 10 proteins to 516 proteins, which together comprised more than 1/5th of the detected astrocyte proteome. Interestingly, numerous critical electron transport chain (ETC) proteins were exclusively detected at the endfeet, while enzymes involved in glycogen storage were mostly distributed to the somata, indicating subcellular metabolic compartmentalization. The endfoot proteome also included numerous proteins that, while previously known to have important contributions to BBB development and function, were not previously known to localize to the endfoot.

Materials AND methods

Isolation of astrocyte endfeet and somata

Animal experiments were approved by the Institutional Animal Care and Use Committee of the University of Maryland and were performed in accordance with the NIH Guide for the Care and Use of Laboratory Animals. Experiments are reported in compliance with the ARRIVE guidelines. Approximately 50 mice were utilized for this study.

Astrocyte endfeet were isolated from adult (22–25 g) male C57BL/6 mice (Figure 1(a)). Prior to brain harvest, mice were euthanized with pentobarbital overdose, whereupon they were intracardially perfused with cold saline and the whole brain was dissected from the skull. All subsequent steps were conducted on ice. First, cerebral microvessels were isolated from mouse brain using the method described by Boulay et al.^28,29 Importantly, these microvessels remain coated with β-dystroglycan and aquaporin-4 positive astrocyte endfeet that are “amputated” from their parent astrocyte somata (Figure 1(b) to (d)). Isolated microvessels were then dissociated into a single cell suspension for 60 minutes at 37 °C in 20 U/mL DNAseI and 100 µg/mL Liberase DL (cat. 5,40,11,60,001 Millipore Sigma), followed by centrifugation at 300 g for 30 min to clear large cells and debris. The remaining suspension was pelleted at 13,000 g for 30 min, and then resuspended in AstroMACS separation buffer (cat. 130-117-336, Miltenyi Biotec. The Miltenyi Biotec adult mouse astrocyte isolation protocol (cat. 130-097-678, Miltenyi Biotec),^30,31 which relies on the plasmalemma marker ACSA-2 to separate astrocytes, was used to isolate astrocyte endfeet from the suspension. Approximately 5 µg of endfoot protein was obtained from a single adult mouse brain.

Figure 1.

Isolation of perivascular astrocyte endfeet. (a) Diagram of the endfoot isolation protocol showing steps of isolation: (i) isolation of microvessels, (ii) proteolytic dissociation of microvessels into single-cell suspension and (iii) column-based magnetic bead separation of endfeet from suspension. (b–d) Micrographs of isolated mouse microvessels (MV) (b–d) immunolabeled for endfoot markers β-dystroglycan (β-DG) and aquaporin-4 (AQP4), and endothelial marker zonulin-1 (ZO-1); DAPI labeling in blue. (e) Micrographs of isolated intact MV, partially digested MV, and isolated endfeet immunolabeled for AQP4 and ZO-1; individual endfeet (bottom row) also exhibit labeling for the mitochondrial marker TOM20, and the astrocyte-specific marker ASCA2; DAPI labeling in blue. (f) Immunoblot for AQP4, ZO-1, the nuclei marker nuclear pore complex (NPC), and the pericyte marker PDGFRβ, and capillary electrophoresis immunoassay for β-DG and CD31 in isolated MV, isolated endfeet, and the column flow-through, showing that MV exhibit all markers, isolated endfeet exhibit AQP4 and β-DG, and the flow-through exhibits ZO-1, CD31, and NPC, indicating enrichment of endfeet and depletion of contaminants. (g) qPCR analysis of isolated astrocyte somata demonstrating expression of Gfap, but not Ng2, Zo-1, Iba1, Mbp, or NeuN mRNA, indicating enrichment of isolated somata. (h) Immunoblot for β-DG in whole brain, isolated astrocyte somata, and isolated endfeet, showing expression of the β-DG endfoot marker in whole brain and endfeet, but not somata.

Astrocyte somata were isolated from adult (22-25 g) male C57BL/6 mice with the Miltenyi Biotec brain dissociation and astrocyte isolation kits (cat. 130-097-678, Miltenyi Biotec).^30,31 Mouse brain endothelial cells were isolated from mouse brain with the Miltenyi Biotec brain dissociation and endothelium isolation kits (cat. 130-097-418; cat. 130-052-301, Miltenyi Biotec). Isolated astrocytes, endothelium, and astrocyte endfeet were immediately frozen in liquid nitrogen or plated and immunolabeled for subsequent experiments.

For somata, a single replicate in the proteomics analysis represented astrocyte somata obtained from a single animal. For endfeet, isolate from 5 animals were pooled together for each replicate. For qPCR, pooling was not necessary.

Conventional immunoblot and capillary electrophoresis immunoassay

For immunoblot assay, tissue samples were lysed in ice-cold lysis buffer (1% Triton X-100 in DPBS) with protease inhibitor (cat. 11,69,74,98,001 Roche). Immunoblots were prepared with standard methods, probed with antibodies raised against-ZO-1 (cat. 40-2200, Invitrogen), PDGFRβ (cat. 3169, Cell Signaling Technology), nuclear pore complex (cat. NBP1-05390, Novus Biologicals), AQP4 (cat. AB3594, Millipore), and visualized with chemiluminescence. 10 µg of total protein was loaded into each well.

The capillary electrophoresis-based Jess system (ProteinSimple) was also used for immunodetection. Protein lysates were mixed with NuPAGE LDS buffer (cat. NP0007, Invitrogen), reducing agent (cat. NP0009, Invitrogen), and the SimpleWes Fluorescent Master Mix (cat. PS-FL01-8, ProteinSimple). 3 µg of total protein per capillary was separated using a standard protocol with a 15-second loading step and a 30-minute separation step. After blocking (cat. 042-203, ProteinSimple), capillaries were incubated with primary antibodies against β-actin (cat. 3700; Cell Signaling Technologies), COX5A (cat. ab180129; Abcam), NDUFA1 (cat. ab195807; Abcam), RECK (cat. ab238162; Abcam), PODXL (cat. ab205350; Abcam), or GABA Transporter 2 (cat. ab229815; Abcam). Data visualization and analysis were performed using the Compass for SW software (version 4.1.0; ProteinSimple).

Quantitative Real-Time polymerase chain reaction

Quantitative real time polymerase chain reaction (qPCR) was used to validate astrocyte somata and endfoot enriched proteins using methods described previously.³² Supplementary Table 1 contains the list of primers.

Proteomics by nano-flow UPLC-coupled high resolution mass spectrometry

Astrocyte endfeet or somata (n = 4 replicates) were solubilized in 5% sodium deoxycholate and were washed, reduced, alkylated and trypsinolyzed in filter as described.^33,34 Tryptic peptides were separated on a nanoACQUITY UPLC analytical column (BEH130 C18, 1.7 μm, 75 μm × 200 mm, Waters) over a 165-minute linear acetonitrile gradient (3–40%) with 0.1% formic acid on a Waters nano-ACQUITY UPLC system and analyzed on a coupled Thermo Scientific Orbitrap Fusion Lumos Tribrid mass spectrometer as described.³⁵ Full scans were acquired at a resolution of 240,000, and precursors were selected for fragmentation by collision-induced dissociation (normalized collision energy at 35%) for a maximum 3-second cycle. Tandem mass spectra were searched against a UniProt mouse reference proteome using the Sequest HT and the MS Amanda algorithms with a maximum precursor mass error tolerance of 10 ppm.^36,37 Carbamidomethylation of cysteine and deamidation of asparagine and glutamine were treated as static and dynamic modifications, respectively. Resulting hits were validated at a maximum false discovery rate (FDR) of 0.01 using the semi-supervised machine learning algorithm Percolator.³⁸ Label-free quantifications were performed using Minora, an aligned Accurate Mass and Retention Time (AMRT) cluster quantification algorithm (Thermo Scientific, 2017). Protein abundance ratios between the endfeet and the bodies were measured by comparing the MS1 peak volumes of peptide ions, which represent total protein abundance, and whose identities were confirmed by MS2 sequencing as described above.

Proteome informatic and statistical analysis

Proteins showing a z-score of fold change >1 with an FDR adjusted ANOVA p-value <0.05 were considered significantly changed and used for further analysis.³⁹ Proteome informatic analysis was performed as described previously.⁴⁰ Ingenuity pathway analysis was used to predict canonical pathways according to the proteins that were significantly different using a Fisher's exact test p-value < 0.01 with a non-zero absolute activity z-score.⁴¹

Immunolabeling

Immunolabeling of isolated vessels or coronal frozen sections of mouse brain was performed as previously described with primary antibodies against GFAP (cat. C9205; Sigma-Aldrich), β-dystroglycan (cat. B-DG-CE; Leica Biosystems), SFXN5 (cat. ab172971; Abcam), ZO-1 (cat. 40-2200; Invitrogen), aquaporin-4 (cat. AB3594; Millipore), TOM-20 (cat. 6,12,278 BD Biosciences), and ASCA2 (cat. 130-099-138; Miltenyi Biotec).⁴² Immunolabelled samples were imaged with standard epifluorescence microscopy (Nikon Eclipse 90i, Nikon Instruments). Negative controls consisted of the secondary antibody applied in the absence of a primary antibody. SFXN5 specificity to the astrocyte mitochondria was also confirmed by demonstrating lack of co-localization with microglia (Supplementary Figure 1).

For quantification of mitochondrial subcellular coverage, images were obtained of mouse cortex from four mice. Either β-dystroglycan (representing endfeet) or GFAP (representing somata and main processes) signals were thresholded at 1.5× background intensity and were used to mask the SFXN5 signals (representing astrocyte mitochondria). Mitochondrial percent area was then calculated by dividing the area occupied by mitochondria by the total area of the mask.

Statistics

Data are presented as mean ± SD. Analysis was performed with Origin Pro (V8; OriginLab Corp, North Hampton MA). Data normality was confirmed with the Shapiro-Wilk test. Student t-test and 1-way ANOVA with Tukey’s HSD post-hoc comparisons were used where appropriate.

Results

Isolation of astrocyte endfeet

To characterize the endfoot proteome, we developed a method to isolate astrocyte endfeet from mouse brain tissue (Figure 1(a)). First, microvessels were prepared from whole brain using previously published methods. Isolated microvessels retain their endfoot layer, with the endfeet becoming “amputated” from their parent somata during isolation (Figure 1(b) to (d)). Isolated microvessels were then dissociated into a suspension of individual vesicles, whereupon endfeet were recovered using the astrocyte isolation kit from Miltenyi Biotec (Figure 1(e)). Isolated endfeet consisted of ∼5–10 µm non-nucleated vesicles, which exhibited immunolabeling for the endfoot marker aquaporin-4 and ASCA-2, the astrocyte specific surface protein (Figure 1(e)). As expected, isolated endfeet also contained numerous TOM20-positive mitochondria (Figure 1(e)).

The sequential steps in the endfoot isolation were validated with immunoassay of cell-specific markers to demonstrate enrichment of the endfoot fraction and depletion of other components of the microvasculature. On immunoassay, endfeet exhibited aquaporin-4 and β-dystroglycan (astrocyte markers), but not zonulin-1 or CD31 (endothelial markers), PDGFRβ (pericyte marker), or nuclear pore complex (nuclei marker) (Figure 1(f)). Immunolabeling experiments showed scant, but overall low levels of zonulin-1 labeling in isolated endfeet (Figure 1(e)). Overall, these experiments indicate low levels of contaminants.

To serve as a comparator, astrocyte somata also were isolated from adult mouse brains using the Miltenyi Biotec brain dissociation and astrocyte isolation kits. qPCR analysis verified that isolated astrocytes exhibited GFAP, but not NeuN, Iba1, MBP, NG2, or zonulin-1 transcripts (Figure 1(g)).

To determine if isolated astrocyte somata retained their endfoot processes, we used immunoblot to probe somata and endfeet for β-dystroglycan, an endfoot marker. β-dystroglycan was detected in endfeet but not somata, indicating that during isolation, mechanical severing and loss of endfoot processes occurs (Figure 1(h)).

The astrocyte endfoot proteome

We submitted four samples of endfeet and somata to nano-flow UPLC coupled high resolution mass spectrometry. A total of 2560 proteins were detected among all samples. Hierarchical clustering indicated markedly different expression patterns between somata and endfeet (Figure 2(a)). Notably, 516 proteins, which comprised more than 1/5th of all detected proteins, were exclusively found in astrocyte endfeet (Figure 2(b)). With regards to the remaining proteins, 41 were detected exclusively in the astrocyte somata. 105 proteins were detected in both the endfoot and the somata, but were significantly more enriched in the somata compared to the endfoot. The remaining proteins were present in the endfoot and somata at similar levels of abundance. Complete proteomics results can be found in supplementary material (Supplementary Table 2).

Figure 2.

Comparative proteomics of the astrocyte endfoot versus somata. (a) Hierarchical clustering of the 2560 proteins detected between somata and endfeet, with red indicating upregulation and values displayed on a log₁₀ scale. (b) Venn diagram of endfoot specific, shared and somata specific. (c) Capillary electrophoresis immunoassay of COX5A, NUFA1, and the loading control actin in whole brain, astrocyte somata, and astrocyte endfeet, confirming proteomics data that showed enrichment of COX5A in endfeet, and of NDUFA1 in somata.

Canonical pathway analysis was performed to investigate patterns of protein expression. Pathways enriched in the endfoot included oxidative phosphorylation, tRNA charging, Rho GTPase signaling, and coagulation system (Table 1). Interestingly, the endfeet also exhibited greater expression of proteins associated with mitochondrial calcium uptake, including MCUR1, MCU, MICU1 and 2, SLC25A23, and IMMT. The astrocyte somata were enriched in glycogen degradation and sirtuin signaling (Table 1).

Table 1.

Pathway analysis of proteomic data.

Pathway	Constituent proteins	P-value	Z-score
Oxidative phosphorylation	COX15,COX5A,COX7A1,DMAC2L,MT-CO3,MT-ND2,MT-ND4L,NDUFA1,NDUFB8,NDUFS6	0.0013	2.53
tRNA charging	DARS2,HARS2,KARS,LARS2,MARS,PARS2,SARS2,WARS	0.000014	2.83
Rho signaling	CDH1,CDH5,EZR,GFAP,GNAZ,GNG3,GNG4,GNG7,PIP4K2B,PIP5K1C,ROCK2,SEPT3,SEPT6,STMN1,VIM,WASF1	0.0023	1.73
Coagulation	F3,FGA,FGB,SERPINA1,SERPINC1	0.0033	1.34
Mitochondrial calcium uptake	MCUR1, MCU, MICU1, MICU2, SLC25A23, IMMT	N/A	N/A
Sirtuin signaling	CDH1,GSK3B,H1F0,Hist1h1e,MT-ND2,MT-ND4L,NAMPT,NDUFA1,NDUFA4L2,NDUFAF2,NDUFB8,NDUFS6,NOS3,POLR1A,SOD3,TIMM13,TIMM17A,TOMM7	0.0025	–0.535
Glycogen degradation	AGL,PGM1,PYGB,PYGM	0.00029	–2

To validate the proteomics data, we performed capillary electrophoresis immunoassay of several oxidative phosphorylation proteins found to differ between cellular compartments. As predicted by the proteomics data, COX5A was enriched at the endfeet, while NDUFA1 was enriched in the somata (Figure 2(c)).

Proteins involved in oxidative phosphorylation are enriched at astrocyte endfeet

To visualize differences in abundance of proteins involved in metabolism, we diagramed the glycolytic, glycogen, glutamate, tricarboxylic acid cycle and ETC pathways (Figure 3). Interestingly, all steps of the glycolytic pathway and tricarboxylic acid cycle were fully expressed by both the endfoot and somata. However, several key ETC subunits were only detectable in endfeet. These included the complex I proteins MTND2, ND4L, NDUFB8, NDUFS6, NDUFAF2 and NDUFA4L2, the complex IV proteins COX15, COX5A, COX7A1, MT-CO3, COA3, COA7, GM11273, and the ATP synthase protein COX5A (Figure 3).

Figure 3.

Differential expression of metabolic pathways in endfoot versus somata. Diagram of the glycolytic pathway (green highlight), tricarboxylic acid cycle (purple highlight), electron transport chain (yellow highlight), and glycogen and glutamate pathways; protein abbreviations are depicted for each metabolic step, and colored tags depict differences in abundance between endfeet and somata, per the key in the upper right; if no difference was reported between endfeet and somata, an encircled B is depicted beside the protein name.

Interestingly, several other metabolic pathways were differentially expressed. Enzymes that are related to glycogen synthesis and degradation were enriched in the somata, while GSK3B, which phosphorylates and inactivates glycogen synthase, was exclusively found in the endfeet (Figure 3). Together, these differences suggest that glycogen dynamics may be localized to the somata. In addition, several differences in glutamate processing were apparent. GAD2, which processes GABA from glutamate was found only in endfeet, whereas GLU1, which processes glutamine from glutamate, was enriched in the somata (Figure 3).

Astrocyte endfeet are densely packed with mitochondria

Since the proteomics data indicated compartmentalization of oxidative phosphorylation within the astrocyte endfeet, we investigated the distribution of mitochondria in astrocytes in vivo. Naïve mouse brain tissues were immunolabeled for sideroflexin-5 (SFXN5). SFXN5 is a mitochondrial amino acid transporter that is capable of transporting serine and potentially citrate, although its function has not been extensively characterized.⁴³ Prior studies have shown that it is exclusively expressed by astrocytic mitochondria, making it a useful marker protein.⁴⁴ In the cortical grey matter, SFXN5+ mitochondria colocalized with GFAP+ astrocytic somata and main branches (Figure 4(a)). Interestingly, SFXN5 signal was increased at the perivascular astrocyte endfeet (Figure 4(b)). Quantification of SFXN5 percent area in the various astrocyte sub-compartments revealed that SFXN5 density was lowest in the background parenchyma (3.46 ± 0.37% area), intermediate in somata (20.72 ± 2.79% area), and highest in endfeet (29.03 ± 1.89% area) (Figure 4(c)).

Figure 4.

Enrichment of mitochondria at the astrocyte endfoot (a, b) Micrographs of mouse brain immunolabeled with SFXN5 (red) and GFAP or β-dystroglycan (β-DG) (green) with DAPI counterstain (blue) showing enrichment of SFXN5-positive astrocyte mitochondria at the perivascular endfoot (b) versus the somata (a) (c) Quantification of astrocyte SFXN5+ mitochondrial density in the background neuropil (BG), GFAP-positive somata, and β-DG positive endfeet, showing increased density in endfeet versus somata; n = 4 mice, * p < 0.05 versus BG, and # p < 0.05 versus somata on ANOVA.

Endfoot proteins indicate new endfoot functions

To better understand the potential function of the endfoot proteins, we performed literature searches on the most abundant endfoot specific proteins (Table 2). Excepting one protein (bile salt export pump), all proteins had been previously detected at the BBB. Interestingly, only four of these proteins had been previously demonstrated to be expressed by astrocytes. Endfoot enrichment of three proteins listed in Table 2 that were not previously reported to be expressed in astrocytes was confirmed with capillary electrophoresis immunoassay of isolated endfeet (Figure 5). To validate astrocyte expression of proteins in Table 2, astrocytes isolated from mouse brain were probed with qPCR for mRNA for proteins in Table 2. Endothelial cells and flow-through from endfoot isolation served as comparators. mRNA for all proteins in Table 2 were detected in astrocyte somata (Supplementary Table 3). Many of these proteins were previously found to be critical to the normal development and function of the BBB, and their abundant expression at the astrocyte endfoot imparts a new importance to the endfoot.

Table 2.

Top proteins exclusively found in the astrocyte endfoot. Literature searches were conducted to determine if any prior studies showed protein expression at the blood-brain barrier, protein expression by astrocytes, and possible endfoot function.

Gene	Protein Name	BBB expression	BBB function	Astrocyte expression
Slc6a13	GABA transporter 2 (GAT2)	Yes⁹³	• Taurine efflux⁷⁹• GABA transport⁹³	Not reported
Slco1a4	Solute carrier organic anion transporter family member 1A2 (OATP1A4/OATP2)	Yes⁹⁴	• organic anion transport⁹⁵• T₃ and T₄ transport⁹⁶• BBB drug transport⁹⁷	Not reported
Mcam	Cell surface glycoprotein (MUC18/Gicerin/CD146)	Yes⁹⁸	• BBB development⁹⁸• T cell diapedesis⁴⁷	Not reported
Lamc3	Laminin subunit γ-3 (LAMC3)	Yes²⁵	• Capillary basement membrane⁹⁹• Astrocyte migration¹⁰⁰• Angiogenesis¹⁰⁰	Not reported
Col4a1	Collagen α-1(IV) chain	Yes¹⁰¹	• Capillary basement membrane¹⁰¹• Angiogenesis ¹⁰²	Yes¹⁰³
Abcb11	Bile salt export pump (BSEP)	Not reported	• Metabolite and drug export^104,105	Not reported
Reck	Reversion-inducing cysteine-rich protein with Kazal motifs (RECK)	Inferred from function	• Cerebral vascularization and BBB differentiation^48,50	Not reported
Podxl	Podocalyxin	Yes¹⁰⁶	• As in renal podocytes, may maintain endfoot morphology⁴⁶	Not reported
Icam2	Intercellular adhesion molecule 2 (ICAM2/LGP55)	Yes¹⁰⁷	• Leukocyte diapedesis ^108,109	Yes¹¹⁰
Abca9	ATP-binding cassette sub-family A member 9	Yes¹¹¹	• BBB drug transport ¹¹²	Yes ¹¹³
Atp2b2	Calcium-transporting ATPase (PMCA2)	Yes^114,115	• Intracellular Ca²⁺ extrusion ¹¹⁶	Yes ¹¹⁶

Figure 5.

Validation of new localization of BBB proteins to the astrocyte endfoot. Capillary electrophoresis immunoassay of isolated astrocyte somata and isolated astrocyte endfeet probed with antibodies against reversion-inducing cysteine-rich protein with Kazal motifs (RECK), GABA transporter 2 (GAT2), and podocalyxin (PODXL), depicted with actin loading controls, verifying endfoot expression and enrichment predicted by mass spectrometry analysis; n = 2 mice.

Discussion

Using a novel technique to isolate astrocyte endfeet from brain tissue, we studied the endfoot proteome. Our results expand the list of known endfoot-specific proteins from 10 to 516 proteins. Furthermore, our data indicate that astrocytes exhibit substantial metabolic compartmentalization, with apparent endfoot sequestration of numerous critical components of the electron transport chain.

Our analysis yielded a total of 2560 proteins. Of these, 516 proteins were exclusively detected in the endfoot, indicating that a surprisingly large portion of the astrocyte proteome is dedicated to endfoot function. A subgroup of the 10 proteins previously known to be enriched at the endfoot were detected in our analysis. As predicted by prior literature, β-dystroglycan and β2-syntrophin were detected only in the endfoot compartment. Integrin α6, GLUT1, α1-syntrophin, dystrophin, dystrobrevin were detected, and were 3.9-, 4.7-, 2.8-, 2.5-, and 7.8-fold more abundant in the endfoot, although these values did not rise to statistical significance after correction for multiple comparisons. Interestingly, although utrophin and aquaporin-4 were detected, these proteins were not significantly enriched at the endfoot. With regards to aquaporin-4, a widely used endfoot marker, vesicular- and Golgi-associated aquaporin-4 in the somata may equilibrate the protein abundance between somata and endfoot.⁴⁵ Finally, neither limitrin nor β4 integrin were detected. Alternative fractionation techniques may be needed to detect these proteins.

Within the most abundant endfoot proteins, many had known BBB functions, although most were not previously known to localize to the endfoot (Table 2). For example, podocalyxin, a large negatively charged protein is known to coat the renal podocyte endfoot and maintains their morphology.⁴⁶ In addition, the cell surface glycoprotein MUC18 was known to regulate leukocyte diapedesis across the BBB, although an endfoot specific role was not known.⁴⁷ Our results have implications for multiple endfoot roles, which are discussed in the following sections.

BBB differentiation

The astrocyte endfoot is an important mediator of BBB differentiation and maintenance via production of various signaling molecules such as Wnt and SHH.^48,49 Our dataset did not contain many of these factors, perhaps because alternative fractionation techniques may be needed to detect these secreted proteins. However, we found that astrocyte endfeet strongly expressed the reversion-inducing cysteine-rich protein with Kazal motifs (RECK). RECK was previously found to be an important mediator of cerebral vascularization and maintenance of BBB differentiation, although it was not previously known to be expressed by the endfoot.^48,50 We also found that endfeet are enriched in platelet derived growth factor receptor α (PDGFRα), which contributes to BBB breakdown and edema formation after intracerebral hemorrhage.⁵¹

Regulation of vascular tone

Due to their close approximation to synapses and the cerebrovasculature, astrocytes were long suspected to mediate functional hyperemia. Astrocytes respond to neuronal activity with intracellular calcium transients that trigger the release of vasoactive substances, resulting in vasoconstriction or dilatation.^6,52–54 Our data indicated enrichment of mitochondrial calcium uptake machinery at the endfoot, including MCUR1, MCU, MICU1, MICU2, SLC25A23, and IMMT. These mitochondrial transport proteins modulate cytosolic calcium and regulate mitochondrial respiration.⁵⁵ Further work will be needed to determine if these specific proteins modulate the calcium transients involved in vasoregulation, which remains an important yet incompletely understood phenomenon.

Astrocyte endfoot metabolism

Our results indicate that numerous subunits of ETC complex I, IV, and ATP synthase are exclusively found in the endfoot (Figure 3). Many of these proteins, including MTND2,⁵⁶ ND4L,⁵⁷ NDUFB8,⁵⁸ NDUFS6,⁵⁹ NDUFAF2,⁶⁰ COX15,⁶¹ COX5,⁶² COA3,⁶³ COA7,⁶⁴ are critical components of their respective ETC complexes and, if deficient result in severe metabolic syndromes. Until recently, it was assumed that astrocytes were energetically quiescent cells that were mostly reliant upon glycolysis.^65,66 Newer data indicate that astrocytes exhibit full expression and function of the tricarboxylic acid cycle and ETC pathways.⁶⁷ Our results suggest that a large portion of astrocyte oxidative phosphorylation may be localized to the endfeet. Since astrocytes mediate ∼30% of total brain oxidative phosphorylation,^68,69 a large proportion of total brain oxidative capacity may therefore take place in the astrocyte endfoot.

Sequestration of ETC enzymes to the endfoot was associated with greater SFXN5+ mitochondrial density at this compartment. In astrocytes, mitochondria are trafficked to and concentrated in areas of high metabolic demand.⁷⁰ Mitochondrial enrichment at the endfoot may be supported by local biogenesis and differential trafficking. In neurons, localized biogenesis was recently reported to support the mitochondrial pool present in axonal neurites.⁷¹ Astrocytes exhibit local mRNA translation at the astrocyte endfoot, which may contribute to compartmentalized mitochondrial biogenesis.²⁹ In perisynaptic astrocyte processes, Miro1 was found to mediate repositioning of mitochondria to areas of active calcium signaling, potentially to support mitochondrial calcium buffering.⁷² In our data, Miro1 was detected in both somata and endfeet, and was ∼2.5-fold more abundant in the endfoot, although this difference was not statistically significant.

The purpose of endfoot metabolic compartmentalization remains unclear. ETC compartmentalization to the endfoot may support its many energy-intensive functions. In other cells, subcellular metabolic compartmentalization is employed to optimize ATP delivery to distant compartments since diffusion of ATP through cytosol is limited by cytoplasmic viscosity, bulky intracellular machinery and labyrinthine ramifications.^73,74 Neurons utilize metabolic compartmentalization, with synapses exhibiting differences in ETC subunit abundance.⁷⁵ Alternatively, compartmentalization of the ETC and mitochondria to the endfoot may support endfoot purinergic signaling.⁷⁶ A highly oxidative endfoot carries implications for CNS disease. After injury, endfeet deteriorate and detach from cerebral vessels which may underlie substantial morbidity.^77,78 Protection of the endfoot after injury should be explored as a potential therapeutic strategy.

Modulation of neuronal signaling

Our results indicate that GABA processing machinery is enriched at the endfoot. These proteins include glutamate decarboxylase 2 (GAD2), which generates GABA from glutamate, and sodium- and chloride-dependent GABA transporter 2 (GAT2), which mediates transmembrane GABA transport. In the brain, GAT2 is present surrounding microvessels,⁷⁹ where it is partially expressed by GABAergic neurons that modulate vascular tone.⁸⁰ Our results suggest that astrocyte endfeet may contribute to perivascular GABAergic signaling, and may modulate resting neuronal tone and neurovascular coupling.^81,82 Given the multiple forms of GAD and GAT expressed by astrocytes and the complex nature of GABAergic signaling in the adult CNS, additional work is needed to determine the precise role of the endfoot in this signaling environment.⁸³

Neuroinflammation

Since the endfeet examined in the present study were obtained from healthy brain tissue, it is unsurprising that many molecules associated with neuroinflammation were not detected. Interestingly, our data indicated endfoot expression of the mitochondrial translocator protein (TSPO). TSPO is a mitochondrial cholesterol transport mediator involved in steroidogenesis and mitochondrial transition pore opening that has attracted great interest as a positron emission tomography (PET) marker for neuroinflammation.^84–86 Reactive astrocytes upregulate TSPO at early time points during neuroinflammation and are important contributors to the brain TSPO PET signal.^87,88 In our dataset, TSPO was detected in both astrocyte somata and endfeet and was 2.7-fold more abundant in the endfoot, although this difference did not rise to statistical significance. Since in some models, astrocytes comprise up to 35% of TSPO positive cells, our results indicate that endfeet may be an important source of TSPO PET signal.⁸⁹ Furthermore, endfoot TSPO may be a potentially important target for TSPO modulation for reduction of neuroinflammation.

Potential limitations

Our study has several potential limitations. The purity of our endfoot proteome rests upon the efficiency of the magnetic cellular separation step, wherein endfeet are separated from endothelial, pericyte, and smooth muscle cell “contaminants.” While data in Figure 1(f) indicate that nucleated contaminants from endothelial cells or pericytes are undetectable with immunoblot or capillary electrophoresis immunoassay, low levels of contaminants can be visually appreciated with immunohistochemistry (Figure 1(e)). While the abundance of these contaminants is low, their presence may theoretically result in false positive endfoot proteins. Future studies of the endfoot proteins reported here should include validation of their endfoot localization.

Our astrocyte proteome of 2560 proteins is smaller than other studies, some of which report up to 7265 proteins using prefractionation techniques.⁹⁰ The “missing” proteins may be less abundant and thus may have remained under the limit of detection or may have become lost or digested during endfoot isolation. In addition, multiple fractionation techniques may be needed to generate a more comprehensive proteome.

There are several other potential limitations. Firstly, we utilized male mice for our experiments. Sex-related differences have been found in the proteome of brain microvessels, particularly in proteins related to mitochondria and the electron transport chain.⁹¹ Future experiments should be conducted in this area. Secondly, it is unclear whether the isolated somata retain other processes, such as perineuronal astrocyte processes. To our knowledge, specific molecular markers or isolation methods for these processes do not exist, which precludes investigation. Thirdly, in our study, isolated endfeet were obtained from whole brain in order to maximize protein yield, which may mask any regional heterogeneity in endfoot composition or energy metabolism. Substantial molecular and functional regional heterogeneity is present in whole astrocytes located in different brain regions,⁹² particularly with regards to endfoot proteins such as AQP4.⁴²

Supplemental Material

sj-zip-1-jcb-10.1177_0271678X211004182 - Supplemental material for A large portion of the astrocyte proteome is dedicated to perivascular endfeet, including critical components of the electron transport chain

Supplemental material, sj-zip-1-jcb-10.1177_0271678X211004182 for A large portion of the astrocyte proteome is dedicated to perivascular endfeet, including critical components of the electron transport chain by Jesse A Stokum, Bosung Shim, Weiliang Huang, Maureen Kane, Jesse A Smith, Volodymyr Gerzanich and J Marc Simard in Journal of Cerebral Blood Flow & Metabolism

Footnotes

Funding

The author(s) disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: This research was supported in part by the University of Maryland School of Pharmacy Mass Spectrometry Center (SOP1841-IQB2014). J.M.S. is supported by grants from the Department of Veterans Affairs (I01BX002889), the Department of Defense (SCI170199), the National Heart, Lung and Blood Institute (R01HL082517) and the NINDS (R01NS060801; R01NS102589; R01NS105633). V.G. is supported by a grant from the NINDS (R01NS107262).

Declaration of conflicting interests

The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Authors’ contributions

JAS conceived of the project, collected and analyzed data, and wrote the manuscript. BS collected data and edited the manuscript. WH and MK collected and analyzed data and edited the manuscript. JAS collected data and edited the manuscript. VG and JMS conceived of the project and edited the manuscript.

ORCID iD

Jesse A Stokum

Supplemental material

Supplemental material for this article is available online.

Acknowledgements

The authors thank the University of Maryland School of Pharmacy Mass Spectrometry Center for their proteomics support.

References

Hertz

Peng

Dienel

GA.

Energy metabolism in astrocytes: high rate of oxidative metabolism and spatiotemporal dependence on glycolysis/glycogenolysis. J Cereb Blood Flow Metab 2007; 27: 219–249.

Azevedo

Carvalho

Grinberg

, et al. Equal numbers of neuronal and nonneuronal cells make the human brain an isometrically scaled-up primate brain. J Comp Neurol 2009; 513: 532–541.

Bushong

Martone

Jones

, et al. Protoplasmic astrocytes in CA1 stratum radiatum occupy separate anatomical domains. J Neurosci 2002; 22: 183–192.

Halassa

Fellin

Takano

, et al. Synaptic islands defined by the territory of a single astrocyte. J Neurosci 2007; 27: 6473–6477.

Nuriya

Yasui

Endfeet serve as diffusion-limited subcellular compartments in astrocytes. J Neurosci 2013; 33: 3692–3698.

Dunn

Hill-Eubanks

Liedtke

, et al. TRPV4 channels stimulate Ca2+-induced Ca2+ release in astrocytic endfeet and amplify neurovascular coupling responses. Proc Natl Acad Sci USA 2013; 110: 6157–6162.

Langer

Gerkau

Derouiche

, et al. Rapid sodium signaling couples glutamate uptake to breakdown of ATP in perivascular astrocyte endfeet. Glia 2017; 65: 293–308.

Derouiche

Pannicke

Haseleu

, et al. Beyond polarity: functional membrane domains in astrocytes and muller cells. Neurochem Res 2012; 37: 2513–2523.

Mathiisen

Lehre

Danbolt

, et al. The perivascular astroglial sheath provides a complete covering of the brain microvessels: an electron microscopic 3D reconstruction. Glia 2010; 58: 1094–1103.

10.

Belanger

Allaman

Magistretti

PJ.

Brain energy metabolism: focus on astrocyte-neuron metabolic cooperation. Cell Metab 2011; 14: 724–738.

11.

Leino

Gerhart

van Bueren

, et al. Ultrastructural localization of GLUT 1 and GLUT 3 glucose transporters in rat brain. J Neurosci Res 1997; 49: 617–626.

12.

Winship

Plaa

Murphy

TH.

Rapid astrocyte calcium signals correlate with neuronal activity and onset of the hemodynamic response in vivo. J Neurosci 2007; 27: 6268–6272.

13.

Mulligan

MacVicar

BA.

Calcium transients in astrocyte endfeet cause cerebrovascular constrictions. Nature 2004; 431: 195–199.

14.

Gordon

Choi

Rungta

, et al. Brain metabolism dictates the polarity of astrocyte control over arterioles. Nature 2008; 456: 745–749.

15.

Girouard

Bonev

Hannah

, et al. Astrocytic endfoot Ca2+ and BK channels determine both arteriolar dilation and constriction. Proc Natl Acad Sci USA 2010; 107: 3811–3816.

16.

Haj-Yasein

Vindedal

Eilert-Olsen

, et al. Glial-conditional deletion of aquaporin-4 (Aqp4) reduces blood-brain water uptake and confers barrier function on perivascular astrocyte endfeet. Proc Natl Acad Sci USA 2011; 108: 17815–17820.

17.

Hayashi

Nomura

Yamagishi

, et al. Induction of various blood-brain barrier properties in non-neural endothelial cells by close apposition to co-cultured astrocytes. Glia 1997; 19: 13–26.

18.

Iliff

Wang

Liao

, et al. A paravascular pathway facilitates CSF flow through the brain parenchyma and the clearance of interstitial solutes, including amyloid beta. Sci Transl Med 2012; 4: 147ra111–147ra108.

19.

Nielsen

Nagelhus

Amiry-Moghaddam

, et al. Specialized membrane domains for water transport in glial cells: high-resolution immunogold cytochemistry of aquaporin-4 in rat brain. J Neurosci 1997; 17: 171–180.

20.

Kalman

Oszwald

Adorjan

Appearance of beta-dystroglycan precedes the formation of glio-vascular end-feet in developing rat brain. Eur J Histochem 2018; 62: 2908–2906.

21.

Inoue

Wakayama

Liu

, et al. Ultrastructural localization of aquaporin 4 and alpha1-syntrophin in the vascular feet of brain astrocytes. Tohoku J Exp Med 2002; 197: 87–93.

22.

Robel

Mori

Zoubaa

, et al. Conditional deletion of beta1-integrin in astroglia causes partial reactive gliosis. Glia 2009; 57: 1630–1647.

23.

Ueda

Baba

Terada

, et al. Immunolocalization of dystrobrevin in the astrocytic endfeet and endothelial cells in the rat cerebellum. Neurosci Lett 2000; 283: 121–124.

24.

Yonezawa

Ohtsuka

Yoshitaka

, et al. Limitrin, a novel immunoglobulin superfamily protein localized to glia limitans formed by astrocyte endfeet. Glia 2003; 44: 190–204.

25.

Radner

Banos

Bachay

, et al. beta2 and gamma3 laminins are critical cortical basement membrane components: ablation of Lamb2 and Lamc3 genes disrupts cortical lamination and produces dysplasia. Dev Neurobiol 2013; 73: 209–229.

26.

Morgello

Uson

Schwartz

, et al. The human blood-brain barrier glucose transporter (GLUT1) is a glucose transporter of gray matter astrocytes. Glia 1995; 14: 43–54.

27.

Nagy

Patel

Ochalski

, et al. Connexin30 in rodent, cat and human brain: selective expression in gray matter astrocytes, co-localization with connexin43 at gap junctions and late developmental appearance. Neuroscience 1999; 88: 447–468.

28.

Boulay

Saubamea

Decleves

, et al. Purification of mouse brain vessels. J Vis Exp 2015; 105: e53208.

29.

Boulay

Saubamea

Adam

, et al. Translation in astrocyte distal processes sets molecular heterogeneity at the gliovascular interface. Cell Discov 2017; 3: 17005–17004.

30.

Batiuk

de Vin

Duque

, et al. An immunoaffinity-based method for isolating ultrapure adult astrocytes based on ATP1B2 targeting by the ACSA-2 antibody. J Biol Chem 2017; 292: 8874–8891.

31.

Kantzer

Boutin

Herzig

, et al. Anti-ACSA-2 defines a novel monoclonal antibody for prospective isolation of living neonatal and adult astrocytes. Glia 2017; 65: 990–1004.

32.

Kurland

Gerzanich

Karimy

, et al. The Sur1-Trpm4 channel regulates NOS2 transcription in TLR4-activated microglia. J Neuroinflammation 2016; 13: 130–106.

33.

Wiśniewski

Zougman

Nagaraj

, et al. Universal sample preparation method for proteome analysis. Nat Methods 2009; 6: 359–362.

34.

Erde

Loo

JA.

Enhanced FASP (eFASP) to increase proteome coverage and sample recovery for quantitative proteomic experiments. J Proteome Res 2014; 13: 1885–1895.

35.

Williamson

Edwards

Verano-Braga

, et al. High-performance hybrid orbitrap mass spectrometers for quantitative proteome analysis: observations and implications. Proteomics 2016; 16: 907–914.

36.

Eng

Fischer

Grossmann

, et al. A fast SEQUEST cross correlation algorithm. J Proteome Res 2008; 7: 4598–4602.

37.

Dorfer

Pichler

Stranzl

, et al. MS Amanda, a universal identification algorithm optimized for high accuracy tandem mass spectra. J Proteome Res 2014; 13: 3679–3684.

38.

Kall

Canterbury

Weston

, et al. Semi-supervised learning for peptide identification from shotgun proteomics datasets. Nat Methods 2007; 4: 923–925.

39.

Benjamini

Hochberg

Controlling the false discovery rate: a practical and powerful approach to multiple testing. J R Stat Soc: Series B (Methodological) 1995; 57: 289–300.

40.

Huang

Jones

, et al. Acute proteomic changes in the lung after WTLI in a mouse model: Identification of potential initiating events for delayed effects of acute radiation exposure. Health Phys 2019; 116: 503–515.

41.

Kramer

Green

Pollard

Jr , et al. Causal analysis approaches in ingenuity pathway analysis. Bioinformatics 2014; 30: 523–530.

42.

Stokum

Mehta

Ivanova

, et al. Heterogeneity of aquaporin-4 localization and expression after focal cerebral ischemia underlies differences in white versus grey matter swelling. Acta Neuropathol Commun 2015; 3: 61–10.

43.

Kory

Wyant

Prakash

, et al. SFXN1 is a mitochondrial serine transporter required for one-carbon metabolism. Science 2018; 362: eaat9528.

44.

Fecher

Trovo

Muller

, et al. Cell-type-specific profiling of brain mitochondria reveals functional and molecular diversity. Nat Neurosci 2019; 22: 1731–1742.

45.

Arnspang

Sundbye

Nelson

, et al. Aquaporin-3 and aquaporin-4 are sorted differently and separately in the trans-Golgi network. PLoS One 2013; 8: e73977.

46.

Nielsen

McNagny

KM.

The role of podocalyxin in health and disease. J Am Soc Nephrol 2009; 20: 1669–1676.

47.

Duan

Xing

Luo

, et al. Targeting endothelial CD146 attenuates neuroinflammation by limiting lymphocyte extravasation to the CNS. Sci Rep 2013; 3: 1687–1604.

48.

Cho

Smallwood

Nathans

Reck and Gpr124 are essential receptor cofactors for Wnt7a/Wnt7b-Specific signaling in mammalian CNS angiogenesis and blood-brain barrier regulation. Neuron 2017; 95: 1056–1073.

49.

Liebner

Plate

KH.

Differentiation of the brain vasculature: the answer came blowing by the Wnt. J Angiogenes Res 2010; 2: 1–02.

50.

Ulrich

Carretero-Ortega

Menendez

, et al. Reck enables cerebrovascular development by promoting canonical Wnt signaling. Development 2016; 143: 1055–1003.

51.

Huang

Khatibi

, et al. PDGFR-alpha inhibition preserves blood-brain barrier after intracerebral hemorrhage. Ann Neurol 2011; 70: 920–931.

52.

Zonta

Angulo

Gobbo

, et al. Neuron-to-astrocyte signaling is Central to the dynamic control of brain microcirculation. Nat Neurosci 2003; 6: 43–50.

53.

Takano

Tian

Peng

, et al. Astrocyte-mediated control of cerebral blood flow. Nat Neurosci 2006; 9: 260–267.

54.

Filosa

Bonev

Straub

, et al. Local potassium signaling couples neuronal activity to vasodilation in the brain. Nat Neurosci 2006; 9: 1397–1403.

55.

Kannurpatti

SS.

Mitochondrial calcium homeostasis: Implications for neurovascular and neurometabolic coupling. J Cereb Blood Flow Metab 2017; 37: 381–395.

56.

Zanolini

Potic

Carrara

, et al. Pure myopathy with enlarged mitochondria associated to a new mutation in MTND2 gene. Mol Genet Metab Rep 2017; 10: 24–27.

57.

Cardol

Lapaille

Minet

, et al. ND3 and ND4L subunits of mitochondrial complex I, both nucleus encoded in Chlamydomonas reinhardtii, are required for activity and assembly of the enzyme. Eukaryot Cell 2006; 5: 1460–1467.

58.

Piekutowska-Abramczuk

Assouline

Matakovic

, et al. NDUFB8 mutations cause mitochondrial complex I deficiency in individuals with Leigh-like encephalomyopathy. Am J Hum Genet 2018; 102: 460–467.

59.

Kirby

Salemi

Sugiana

, et al. NDUFS6 mutations are a novel cause of lethal neonatal mitochondrial complex I deficiency. J Clin Invest 2004; 114: 837–845.

60.

Herzer

Koch

Prokisch

, et al. Leigh disease with brainstem involvement in complex I deficiency due to assembly factor NDUFAF2 defect. Neuropediatrics 2010; 41: 30–34.

61.

Antonicka

Mattman

Carlson

, et al. Mutations in COX15 produce a defect in the mitochondrial heme biosynthetic pathway, causing early-onset fatal hypertrophic cardiomyopathy. Am J Hum Genet 2003; 72: 101–114.

62.

Baertling

Al-Murshedi

Sanchez-Caballero

, et al. Mutation in mitochondrial complex IV subunit COX5A causes pulmonary arterial hypertension, lactic acidemia, and failure to thrive. Hum Mutat 2017; 38: 692–703.

63.

Ostergaard

Weraarpachai

Ravn

, et al. Mutations in COA3 cause isolated complex IV deficiency associated with neuropathy, exercise intolerance, obesity, and short stature. J Med Genet 2015; 52: 203–207.

64.

Martinez Lyons

Ardissone

Reyes

, et al. COA7 (C1orf163/RESA1) mutations associated with mitochondrial leukoencephalopathy and cytochrome c oxidase deficiency. J Med Genet 2016; 53: 846–849.

65.

Pellerin

Magistretti

PJ.

Glutamate uptake into astrocytes stimulates aerobic glycolysis: a mechanism coupling neuronal activity to glucose utilization. Proc Natl Acad Sci USA 1994; 91: 10625–10629.

66.

Bouzier-Sore

Pellerin

Unraveling the complex metabolic nature of astrocytes. Front Cell Neurosci 2013; 7: 179.

67.

Lovatt

Sonnewald

Waagepetersen

, et al. The transcriptome and metabolic gene signature of protoplasmic astrocytes in the adult murine cortex. J Neurosci 2007; 27: 12255–12266.

68.

Lebon

Petersen

Cline

, et al. Astroglial contribution to brain energy metabolism in humans revealed by 13C nuclear magnetic resonance spectroscopy: elucidation of the dominant pathway for neurotransmitter glutamate repletion and measurement of astrocytic oxidative metabolism. J Neurosci 2002; 22: 1523–1531.

69.

Berkich

Henry

, et al. Neuroglial metabolism in the awake rat brain: CO2 fixation increases with brain activity. J Neurosci 2004; 24: 11273–11279.

70.

Jackson

O'Donnell

Takano

, et al. Neuronal activity and glutamate uptake decrease mitochondrial mobility in astrocytes and position mitochondria near glutamate transporters. J Neurosci 2014; 34: 1613–1624.

71.

Van Laar

Arnold

Howlett

, et al. Evidence for compartmentalized axonal mitochondrial biogenesis: mitochondrial DNA replication increases in distal axons as an early response to Parkinson's disease-relevant stress. J Neurosci 2018; 38: 7505–7515.

72.

Stephen

Higgs

Sheehan

, et al. Miro1 regulates activity-driven positioning of mitochondria within astrocytic processes apposed to synapses to regulate intracellular calcium signaling. J Neurosci 2015; 35: 15996–16011.

73.

Luby-Phelps

Cytoarchitecture and physical properties of cytoplasm: volume, viscosity, diffusion, intracellular surface area. Int Rev Cytol 2000; 192: 189–221.

74.

Zhou

Rivas

Minton

AP.

Macromolecular crowding and confinement: biochemical, biophysical, and potential physiological consequences. Annu Rev Biophys 2008; 37: 375–397.

75.

Stauch

Purnell

Fox

HS.

Quantitative proteomics of synaptic and nonsynaptic mitochondria: insights for synaptic mitochondrial vulnerability. J Proteome Res 2014; 13: 2620–2636.

76.

Pelligrino

Vetri

HL.

Purinergic mechanisms in gliovascular coupling. Semin Cell Dev Biol 2011; 22: 229–236.

77.

Kneihsl

Niederkorn

Deutschmann

, et al. Increased Middle cerebral artery mean blood flow velocity index after stroke thrombectomy indicates increased risk for intracranial hemorrhage. J NeuroIntervent Surg 2018; 10: 882–887.

78.

Kuroiwa

Shibutani

Okeda

Nonhyperemic blood flow restoration and brain edema in experimental focal cerebral ischemia. J Neurosurg 1989; 70: 73–80.

79.

Zhou

Holmseth

Guo

, et al. Deletion of the gamma-aminobutyric acid transporter 2 (GAT2 and SLC6A13) gene in mice leads to changes in liver and brain taurine contents. J Biol Chem 2012; 287: 35733–35746.

80.

Vaucher

Tong

Cholet

, et al. GABA neurons provide a rich input to microvessels but not nitric oxide neurons in the rat cerebral cortex: a means for direct regulation of local cerebral blood flow. J Comp Neurol 2000; 421: 161–171.

81.

Yoon

Lee

CJ.

GABA as a rising gliotransmitter. Front Neural Circuits 2014; 8: 141.

82.

Fergus

Lee

KS.

GABAergic regulation of cerebral microvascular tone in the rat. J Cereb Blood Flow Metab 1997; 17: 992–1003.

83.

Ishibashi

Egawa

Fukuda

Diverse actions of astrocytes in GABAergic signaling. IJMS 2019; 20: 2964.

84.

Chung

Chen

Midzak

, et al. Drug ligand-induced activation of translocator protein (TSPO) stimulates steroid production by aged brown Norway rat leydig cells. Endocrinology 2013; 154: 2156–2165.

85.

McEnery

Snowman

Trifiletti

, et al. Isolation of the mitochondrial benzodiazepine receptor: association with the voltage-dependent anion channel and the adenine nucleotide carrier. Proc Natl Acad Sci USA 1992; 89: 3170–3174.

86.

Lee

Park

Nam

, et al. Translocator protein (TSPO): the new story of the old protein in neuroinflammation. BMB Rep 2020; 53: 20–27.

87.

Tournier

Tsartsalis

Ceyzeriat

, et al. Astrocytic TSPO upregulation appears before microglial TSPO in Alzheimer's disease. J Alzheimers Dis 2020; 77: 1043–1056.

88.

Lavisse

Guillermier

Herard

, et al. Reactive astrocytes overexpress TSPO and are detected by TSPO positron emission tomography imaging. J Neurosci 2012; 32: 10809–10818.

89.

Pannell

Economopoulos

Wilson

, et al. Imaging of translocator protein upregulation is selective for pro-inflammatory polarized astrocytes and microglia. Glia 2020; 68: 280–297.

90.

Han

Jin

Woo

, et al. Proteomic analysis of mouse astrocytes and their secretome by a combination of FASP and StageTip-based, high pH, reversed-phase fractionation. Proteomics 2014; 14: 1604–1609.

91.

Cikic

Chandra

Harman

, et al. Sexual differences in mitochondrial and related proteins in rat cerebral microvessels: a proteomic approach. J Cereb Blood Flow Metab 2020; 41: 897–412.

92.

Bayraktar

Fuentealba

Alvarez-Buylla

, et al. Astrocyte development and heterogeneity. Cold Spring Harb Perspect Biol 2014; 7: a020362.

93.

Takanaga

Ohtsuki

Hosoya

, et al. GAT2/BGT-1 as a system responsible for the transport of gamma-aminobutyric acid at the mouse blood-brain barrier. J Cereb Blood Flow Metab 2001; 21: 1232–1239.

94.

Gao

Stieger

Noe

, et al. Localization of the organic anion transporting polypeptide 2 (Oatp2) in capillary endothelium and choroid plexus epithelium of rat brain. J Histochem Cytochem 1999; 47: 1255–1264.

95.

Yaguchi

Tachikawa

Zhang

, et al. Organic anion-transporting polypeptide 1a4 (Oatp1a4/Slco1a4) at the blood-arachnoid barrier is the major pathway of sulforhodamine-101 clearance from cerebrospinal fluid of rats. Mol Pharm 2019; 16: 2021–2027.

96.

Abe

Kakyo

Sakagami

, et al. Molecular characterization and tissue distribution of a new organic anion transporter subtype (oatp3) that transports thyroid hormones and taurocholate and comparison with oatp2. J Biol Chem 1998; 273: 22395–22401.

97.

Ose

Kusuhara

Endo

, et al. Functional characterization of mouse organic anion transporting peptide 1a4 in the uptake and efflux of drugs across the blood-brain barrier. Drug Metab Dispos 2010; 38: 168–176.

98.

Chen

Luo

Hui

, et al. CD146 coordinates brain endothelial cell-pericyte communication for blood-brain barrier development. Proc Natl Acad Sci USA 2017; 114: E7622–E7631.

99.

Radner

French

, et al. The gamma3 chain of laminin is widely but differentially expressed in murine basement membranes: expression and functional studies. Matrix Biol 2012; 31: 120–134.

100.

Gnanaguru

Bachay

Biswas

, et al. Laminins containing the beta2 and gamma3 chains regulate astrocyte migration and angiogenesis in the retina. Development 2013; 140: 2050–2060.

101.

Armulik

Genove

Mae

, et al. Pericytes regulate the blood-brain barrier. Nature 2010; 468: 557–561.

102.

Poschl

Schlotzer-Schrehardt

Brachvogel

, et al. Collagen IV is essential for basement membrane stability but dispensable for initiation of its assembly during early development. Development 2004; 131: 1619–1628.

103.

Liesi

Kauppila

Induction of type IV collagen and other basement-membrane-associated proteins after spinal cord injury of the adult rat may participate in formation of the glial scar. Exp Neurol 2002; 173: 31–45.

104.

Schinkel

Smit

van Tellingen

, et al. Disruption of the mouse mdr1a P-glycoprotein gene leads to a deficiency in the blood-brain barrier and to increased sensitivity to drugs. Cell 1994; 77: 491–502.

105.

Mahringer

Fricker

ABC transporters at the blood-brain barrier. Expert Opin Drug Metab Toxicol 2016; 12: 499–508.

106.

Agarwal

Lippmann

Shusta

EV.

Identification and expression profiling of blood-brain barrier membrane proteins. J Neurochem 2010; 112: 625–635.

107.

Daneman

Zhou

Agalliu

, et al. The mouse blood-brain barrier transcriptome: a new resource for understanding the development and function of brain endothelial cells. PLoS One 2010; 5: e13741.

108.

Steiner

Coisne

Cecchelli

, et al. Differential roles for endothelial ICAM-1, ICAM-2, and VCAM-1 in shear-resistant T cell arrest, polarization, and directed crawling on blood-brain barrier endothelium. J Immunol 2010; 185: 4846–4855.

109.

Gorina

Lyck

Vestweber

, et al. beta2 integrin-mediated crawling on endothelial ICAM-1 and ICAM-2 is a prerequisite for transcellular neutrophil diapedesis across the inflamed blood-brain barrier. J Immunol 2014; 192: 324–337.

110.

Satoh

Kim

Kastrukoff

, et al. Expression and induction of intercellular adhesion molecules (ICAMs) and major histocompatibility complex (MHC) antigens on cultured murine oligodendrocytes and astrocytes. J Neurosci Res 1991; 29: 1–12.

111.

Uchida

Ohtsuki

Katsukura

, et al. Quantitative targeted absolute proteomics of human blood-brain barrier transporters and receptors. J Neurochem 2011; 117: 333–345.

112.

Shen

Zhang

ABC transporters and drug efflux at the blood-brain barrier. Rev Neurosci 2010; 21: 29–53.

113.

Sengupta

Mense

Lan

, et al. Gene expression profiling of human primary astrocytes exposed to manganese chloride indicates selective effects on several functions of the cells. Neurotoxicology 2007; 28: 478–489.

114.

Badhwar

Stanimirovic

Hamel

, et al. The proteome of mouse cerebral arteries. J Cereb Blood Flow Metab 2014; 34: 1033–1046.

115.

Guo

Zhou

Xing

, et al. The vasculome of the mouse brain. PLoS One 2012; 7: e52665.

116.

Fresu

Dehpour

Genazzani

, et al. Plasma membrane calcium ATPase isoforms in astrocytes. Glia 1999; 28: 150–155.

Supplementary Material

Please find the following supplemental material available below.

For Open Access articles published under a Creative Commons License, all supplemental material carries the same license as the article it is associated with.

For non-Open Access articles published, all supplemental material carries a non-exclusive license, and permission requests for re-use of supplemental material or any part of supplemental material shall be sent directly to the copyright owner as specified in the copyright notice associated with the article.

3.31 MB

0.00 MB