Age-dependent Decline of Copper Clearance at the Blood-Cerebrospinal Fluid Barrier

Abstract

The homeostasis of copper (Cu) in the central nervous system is regulated by the blood-brain barrier and blood-cerebrospinal (CSF) barrier (BCB) in the choroid plexus. While proteins responsible for Cu uptake, release, storage and intracellular trafficking exist in the choroid plexus, the influence of age on Cu clearance from the CSF via the choroid plexus and how Cu transporting proteins contribute to the process are unelucidated. This study was designed to test the hypothesis that the aging process diminishes Cu clearance from the CSF of rats by disrupting Cu transporting proteins in the choroid plexus. Data from ventriculo-cisternal perfusion experiments demonstrated greater ⁶⁴Cu radioactivity in the CSF effluents of older rats (18 months) compared to younger (1 month) and adult (2 months) rats, suggesting much slower removal of Cu by the choroid plexus in old animals. Studies utilizing qPCR and immunofluorescence revealed an age-specific expression pattern of Cu transporting proteins in the choroid plexus. Moreover, proteomic analyses unraveled age-specific proteomes in the choroid plexus with distinct pathway differences, particularly associated with extracellular matrix and neurodevelopment between young and old animals. Taken together, these findings support an age-dependent deterioration in CSF Cu clearance, which appears to be associated with altered subcellular distribution of Cu transporting proteins and proteomes in the choroid plexus.

1. Introduction

Copper (Cu) is an essential element to normal brain function. Two brain barrier systems, i.e., the blood-brain barrier (BBB) and blood-cerebrospinal fluid (CSF) barrier (BCB) are crucial in maintaining Cu homeostasis in the central nervous system. Reports in literature have established a relationship between Cu dyshomeostasis and numerous neurodegenerative diseases. For example, Cu dyshomeostasis has been suggested to be a risk factor in Alzheimer’s disease (AD), although the controversy remains as to whether Cu-overload or Cu-deficiency ultimately leads to pathogenesis (Bagheri et al. 2018). Compelling evidence also shows Cu ions participate in aggregation of β-amyloid (Aβ) to form amyloid plaques, the hallmark of AD (Atwood et al. 2004; Miller et al. 2006). Recent findings by our group further suggest Cu may act as a “switch” metal in adult neurogenesis in the brain subventricular zone, underlying the nonmotor syndrome of Parkinsonian disorder (Fu et al. 2016; Winner et al. 2011) and olfaction (Adamson et al., 2021). Nonetheless, the precise mechanism by which brain barrier systems regulate the Cu homeostasis in the brain remains elusive.

The functional basis of Cu transport at brain barriers pertain to a group of Cu transport proteins, including high-affinity Cu transport protein-1 (CTR1), divalent metal transporter-1 (DMT1), ATPase copper transporting alpha (ATP7A) and ATPase copper transporting beta (ATP7B) (Zheng and Monnot 2012). Among these, ATP7A and ATP7B are well characterized in their biological functions by disease phenotypes and research models. For instance, defective hepatic ATP7B can lead to failed Cu excretion into the bile, causing excessive hepatic Cu accumulation, and result in the clinical manifestations of Wilson’s disease. Studies from our group have established that the choroid plexus, a predominant regulator of Cu transport between the blood and CSF, possesses various Cu transporters. Exposures to toxic metals or conditions of iron dysregulation (deficiency or overload) greatly influences the expression as well as subcellular distribution of Cu transporters (Fu et al. 2015). Noticeably also, the damage to the brain barrier integrity has been linked to metal-induced brain disorders (Kim et al. 2013; Zheng 2001a). At the BCB, the tight junctions between choroidal epithelial cells are the structural basis of barrier’s permeability. Age-dependent impairment in the BBB structure has been reported (Bors et al. 2018), however, whether or not the aging process influences BCB integrity and tight junctional protein expression is unknown. Further, the question as to how ageing impacts the expression of Cu transporters in the choroid plexus remains unanswered.

The expression and intracellular distribution/trafficking of Cu transporting proteins can be modulated by a variety of factors, including Cu levels, chemical modulation and/or disease status (Fu et al. 2014; Kuo et al. 2006; Monnot et al. 2011; Sechi et al. 2007). Evidence in literature also suggests biological processes such as intracellular protein sorting determine normal Cu transport function (Harris 2000; Jain et al. 2015). Rapid advancement in proteomic technologies has allowed for discovery of cellular pathways in the choroid plexus modulating membrane expression of drug transporters, secretion of bioactive molecules, and maintenance of its metabolic activity, among many others (Lepko et al. 2019; Lun et al. 2015; Redzic and Segal 2004; Sathyanesan et al. 2012; Silva-Vargas et al. 2016; Thouvenot et al. 2006; Uchida et al. 2015; Zheng 2001b). Given the importance of the choroid plexus in controlling Cu movement across the BCB, it becomes imperative to understand age-related alterations in the choroidal proteome.

The current study was designed to test the hypothesis that Cu transport at the BCB, particularly its clearance from the CSF, was age-dependent and relied on the status of critical Cu transporting proteins in the choroid plexus. We employed a well-established ventriculo-cisternal (VC) perfusion technique to estimate Cu clearance at the BCB in young, adult, and old rats. In addition, we investigated the expression patterns of Cu transporters and tight junction genes at the choroid plexus. Furthermore, we used a proteomic technique to explore age-dependent proteome differences. This research provides insights into the role of the BCB in modulating brain Cu homeostasis at different ages.

2. Materials and Methods

2.1. Chemicals and Reagents

Chemicals and reagents were purchased from the following sources: radioactive ⁶⁴Cu (specific activity 15–30 mCi/μg) from Mallinckrodt Institute of Radiology at Washington University (Saint Louis, MO); radioactive ¹⁴C-sucrose (specific activity: 55 mCi/mmol) from Moravek Biochemicals (Brea, CA); CuCl₂ from Sigma Chemicals (St. Louis, MO); Alexa Fluor^™ 488 goat anti-rabbit IgG (H+L), Alexa Fluor^™ 488 goat anti-mouse IgG (H+L), and Alexa Fluor^™ 568 goat anti-chicken IgY (H+L) from Invitrogen (Waltham, MA); Horseradish peroxide (HRP)-conjugated goat anti-rabbit IgG(H+L), HRP-conjugated goat anti-mouse IgG(H+L), and Fluoromount-G™ Slide Mounting Medium from SouthernBiotech (Birmingham, AL); TRIzol Reagent, Pierce™ BCA Protein Assay Kit and Gold Anti-Fade Mountant from Thermo Fisher (Waltham, MA); iScript cDNA Synthesis Kit from Bio-Rad and Clarity ^™ Western ECL Substrate from Bio-Rad (Hercules, CA); Protease Inhibitor Cocktail from MilliporeSigma (Burlington, MA); Precellys24 tissue homogenizer from Bertin Technologies (Rockville, MD); and Min-spin C18 columns from The Nest Group (Southborough, MA). The chemicals and reagents were of analytical grade, HPLC grade or the best available pharmaceutical grade.

2.2. Animals

Sprague-Dawley rats were purchased from Harlan Sprague Dawley Inc. (Indianapolis, IN) and assigned into 3 groups, i.e., young (4-week-old), adult (8-week-old) and old (18-month-old). Upon arrival, rats were housed in a temperature-controlled room under a 12 hr-light/12 hr-dark cycle and had free access to distilled-deionized water and Purina semi-purified rat chow (Purina Mills TestDiet, Richmond, IN). Animals were acclimated for one week prior to experimentation. Animal use was approved by Purdue University Animal Care and Use Committee.

2.3. In situ Ventriculo-Cisternal (VC) Perfusion

The VC perfusion technique has been routinely used in our laboratory to study the kinetics of BCB substance clearance (Fu et al. 2014; Monnot et al. 2011; Shen et al. 2020; Wang et al. 2008). In brief, rats were anesthetized with ketamine/xylazine (75:10 mg/kg, l mg/kg, ip.) and immobilized in a rat stereotaxic apparatus with their head fixed. A midline cutaneous incision was made from the forehead to neck to expose the skull. A burr hole was made in the skull by a drill bit, and the internal cannula was then implanted through the guide cannula into the lateral ventricle based on the following three-dimensional coordinates: 0.8 mm posterior to bregma (X), 1.5 mm lateral to midline (Y), and 3.5 mm vertical (Z) in adult rats; 0.7 mm X, 1.4 mm Y, and 3.3 mm Z in young rats; and 0.8 mm X, 1.6 mm Y, and 3.5 mm Z in old rats.

Pre-gassed artificial CSF (aCSF) containing ⁶⁴Cu (0.4 μM as CuCl₂, 35 μCi/mL) and the space marker ¹⁴C-sucrose (0.5 μCi/mL) were infused into the lateral ventricle at a rate of 28 μL/min by the micropump (Harvard compact infusion pump, model 975, Holliston, MA). A 26G butterfly needle was simultaneously inserted into the cisterna magna to collect CSF outflows during a total of seven 10-min intervals throughout the 70-min perfusion process. The volume of CSF outflows for each sample were determined by weight assuming a CSF density of 1 g/ml. The radioactivity of ⁶⁴Cu and ¹⁴C-sucrose in each sample were counted on a Perkin-Elmer Wizard-1480 Gamma-counter (Shelton, CT) and a Packard Tri-Carb 2900 TR Liquid Scintillation Analyzer (Waltham, MA), respectively. Percentages of recovered radioactivity of ⁶⁴Cu or ¹⁴C-sucrose in CSF effluents were calculated at each time point.

$$\begin{gathered} \%\text{radioactivity}\text{in}\text{CSF}\text{effluent}{=}^{64}{\text{Cu}}_{\text{out}}\text{(in}\text{outflow)}{/}^{64}{\text{Cu}}_{\text{in}}\text{(in}\text{inflow)} \\ {\text{or}\text{[}}^{14}{\text{C]}\text{sucrose}}_{\text{out}}\text{(in}\text{outflow)}\text{/}{[}^{64}{\text{C]}\text{sucrose}}_{\text{in}}(\text{in}\text{inflow}) \end{gathered}$$

The lost radioactivity in the VC perfusion was considered to be removed from the CSF by the BCB efflux mechanism. Recovered radioactivity data of the last five time points were used to calculate the individual steady-state percentage of ⁶⁴Cu and ¹⁴C-sucrose in the CSF effluent. The Cu clearance rate (μL/min), defined as the volume of aCSF cleared of Cu per minute, was calculated using the method by Al-Sarraf et al. (2000):

$$\text{Clearance}(\mathrm{\mu }\text{L}/\text{min})=\frac{{\text{F}}_{\text{in}}\times \left({\text{Cu}}_{\text{in}}-\text{RD}\times {\text{Cu}}_{\text{out}}\right)}{\left({\text{Cu}}_{\text{in}}+{\text{Cu}}_{\text{out}}\right)/2}$$

where Cu_in and Cu_out represent the ⁶⁴Cu disintegrations per minute (dpm) per unit volume (dpm/mL) in the inflow and outflow aCSF, respectively; F_in is the perfusion rate (μL/min), and RD is the % of the Recovered Radioactivity at steady-state for [¹⁴C] sucrose. Fig. 1A illustrates the VC perfusion experiment.

Figure 1. Cu clearance by the choroid plexus using the ventriculo-cisternal perfusion technique.

(A). Graphic illustration of the VC perfusion experiment. The perfusate containing ¹⁴C-sucrose (space marker) and ⁶⁴Cu in aCSF was infused into brain ventricle of rats in various ages, and the aCSF outflow was collected in cisterna magna to estimate the Cu efflux at the BCB. (B-C). The time course of the radioactivity of ⁶⁴Cu and ¹⁴C-sucrose recovered in the CSF outflow during a 70-min VC perfusion period. (D). The steady-state percentages of radioactivity of ¹⁴C-sucrose, ⁶⁴Cu, and ⁶⁴Cu corrected by ¹⁴C-sucrose. (E). Cu clearance rates at BCB of different ages. Data represent mean ± SD, n=4; *: p < 0.05, as compared to the young.

2.4. Quantitative Real-Time RT-PCR (qPCR)

Expressions of mRNAs encoding critical Cu transporters (Ctr1, Atp7a, Atp7b, Dmt1), tight-junction proteins (Cx43, Cldn5, Cldn12, Occludin, Zo-1, Zo-2) and house-keeping protein (Gapdh) in the choroid plexus were quantified by qPCR. Briefly, total RNA was isolated from freshly isolated choroid plexuses using TRIzol Reagent as per manufacturer’s instructions. An aliquot of RNA (1 μg) was reverse-transcribed into cDNA using the BioRad iScript cDNA synthesis kit. The iTaq Universal SYBR Green Supermix was used for qPCR analyses. Amplifications were run by the CFX Connect Real-Time PCR Detection system with an initial 3-min denaturation at 95°C, followed by 40 cycles of 30-s denaturation at 95°C, 10-s gradient 55.0°C-65.0°C, and a 30-s extension at 72°C. A dissociation curve was used to verify that the majority of fluorescence detected could be attributed to the labeling of specific PCR products, and to verify the absence of primer dimers and sample contamination. Each qPCR reaction was run in triplicate. The relative mRNA expression ratios between groups were calculated using 2^−ΔΔCt cycle time method. After confirming that the reference gene was not changed, the cycle time (Ct) values of interested genes were normalized with that of the reference gene in the same sample to obtain ΔΔCt values. The amplification efficiencies of target genes and the internal reference were examined by determining the variations of the cycle time with a series of control template dilutions. The forward and reverse primers for genes of interest in this study were designed using Primer Express 3.0 software and listed in Table 1.

Table 1.

Primers for mRNAs of Research Interest

Genes	Genbank Accession	Forward primer (5’–3’)	Reverse primer (5’–3’)
Cldn5	NM_031701.2	CTGCCTTAATGTCCGGTGGT	AAACCCCAAGCTTCGAGGAG
Cldn12	NM_001100813	TTACCATCCATAGCCTCCAA	TATGCCCAATCTGGAACACT
Ocln	NM_031329	TATGATGAACAGCCCCCTAA	TCGACTCTTTCCGCATAGTC
Zo-1	NM_001106266	GCAGCCAGTTCAAACAAAGT	CGGTCCAAAGATGGTTACAG
Zo-2	NM_053773	TGGAGGGGATGGATGATGAC	CGCCGTCCGTATCTTCAAAG
Cx43	NM_012567	CCTGATGACCTGGAGATTTA	CCTGATGACCTGGAGATTTA
Ctr1	NM_133600	CTAAACGCAGGCCGAGTTTC	TTGGGATGGGCAGGTTCA
Atp7a	NM_052803	CCCTCAACAGCGTCGTCACT	ACTAGCAGCATCCCCAAAGG
Atp7b	NM_012511	GCCTACAAATCGCTGAGACAC	CTCCGCCTTTTCAGCTATGG
Dmt1	NM_001249798.2	CAGTGCTCTGTACGTAACCTGTAAGC	CGCAGAAGAACGAGGACCAA
Gapdh	NM_017008	CCTGGAGAAACCTGCCAAGTAT	AGCCCAGGATGCCCTTTAGT

Note: Gapdh was used a housekeeping gene to calculate the relative expressions of genes of interest across samples.

2.5. Immunofluorescent Staining and Confocal Microscopy

Animals were anesthetized by ketamine/xylazine (75:10 mg/kg, 1 mg/kg, ip.) before the transcardial perfusion. Rats were fixed with the ventral side facing upward on a foam platform. An incision was made on the chest to expose the heart. A butterfly needle was inserted into the left ventricle and advanced for about 1 cm, followed by fixing the needle in the heart by using a hemostat. An incision was then made on the right atrium. A volume of 50 mL ice-cold PBS was infused through the butterfly needle to remove the blood in the brain. After PBS infusion, another volume of 50 mL ice-cold 4% paraformaldehyde (PFA) in PBS was infused through the same needle to fix the brain. A successful transcardial perfusion could be validated by observation of significant tremors shortly after starting the PFA infusion. The blood-free brain was then extracted from the skull and fixed in 20 mL 4% PFA for additional 24 hours at 4°C. After 3 rinses with PBS, the fixed brain was dehydrated in 30% sucrose in PBS at 4°C until it sank (usually within 3 days). The brain samples covering the lateral ventricle posterior to the bregma were cut into 40-μm slices by a microtome (Microm HM 450, Thermo Scientific, Walldorf, Germany).

For IHC analyses, brain slices were washed with PBS 3 times to remove the residual sucrose. The slices were then blocked and permeabilized in 5% normal goat serum in PBST containing 0.3% TritonX-100. The following primary antibody incubation was at 4 °C overnight. Catalog numbers and dilutions of primary antibodies used in this study were listed in Table 2. Primary antibody-treated brain slices were washed 3 times with PBST (each wash taking 5 minutes) and then incubated with corresponding fluorophore-conjugated secondary antibodies (1:500 dilution) at room temperature for 1 hour prevented from light. These slices were then washed 3 times with PBST (each wash taking 5 minutes) and counterstained by DAPI. Stained brain slices were mounted onto Superfrost™ Plus Microscope Slides using Fluoromount-G™ Slide Mounting Medium.

Table 2.

Primary Antibodies and Dilutions Used for IHC and WB.

Target	IHC Dilutions	WB Dilutions	Host Species	Catalog No.
CTR1	1:200	1:1000	Rabbit	Novus NB100–402
DMT1	1:200	1:1000	Rabbit	Proteintech 20507–1-AP
ATP7A	1:200	1:1000	Mouse	Invitrogen MA5–27720
ATP7B	1:200	1:1000	Rabbit	Santa Cruz sc-32445
TTR	1:400	NA	Chicken	Invitrogen PA5–20742
beta-actin	NA	1:5000	Mouse	Sigma-Aldrich A-5316
Collagen IV	1:200	NA	Rabbit	Invitrogen PA5–115036

Note: TTR was used as a marker to identify the CPECs. Beta-actin serves as a loading control for the WB assay.

The confocal images were acquired by Nikon A1Rsi Confocal System under 60x magnification. Transthyretin (TTR) was used as a marker for choroid plexus epithelial cells (CPECs) in this study. Collagen deposition in the choroid plexus was also assessed by measuring the “fluorescence thickness” of the green collagen IV signal using functions embedded in NIS-Element Advanced Research Software (Tokyo, Japan). 30 locations with the collagen lining across multiple brain slices per animal were randomly selected and used to quantify collagen deposition thickness on the CPEC basolateral side or on the choroidal endothelial side, respectively. The 30 thickness measurements per animal were averaged to represent the collagen deposition status on the CPEC basolateral aspect or on the choroidal endothelium. The negative control group was treated with fluorophore-conjugated secondary antibodies only in order to verify the specific binding of the primary antibodies in IHC assays. Images are provided in Supplemental Fig. 1S.

2.6. Western Blot (WB) Analysis

WB analysis was performed as previously described with slight modifications (Liu et al. 2018). Briefly, freshly isolated blood-free choroid plexus tissues were homogenized in RIPA buffer supplemented with a protease inhibitor cocktail. The protein concentrations in choroid plexus lysates were measured by the BCA kit. Upon loading onto the SDS-PAGE gel, equal amounts of protein (20 μg) were used across samples. Electrophoresis was run under 80 V for 15 minutes and 100 V for an hour. The proteins were then transferred onto a methanol-activated PVDF membrane under 100 V for one hour in an ice-filled tank. The membrane was blocked in 0.22-μm filtered 5% BSA in TBS buffer (the block solution) at room temperature for one hour. The membrane was then incubated with corresponding primary antibody diluted by the block solution at 4 °C overnight (Table 2). The loading control used in this study was beta-actin. After three 5-minute washes in TBS, the membrane was probed with HRP-conjugated secondary antibody (dilution 1:3000) for one hour at room temperature. Following sufficient washes by TBS, immunoblotting images were captured using ECL Substrate by BIO-RAD ChemiDoc^™ XRS system. Relative protein expression levels were semi-quantitatively analyzed through the relative optic density (OD) ratios against the beta-actin signal in ImageJ (Bethesda, Maryland, USA).

2.7. Proteomic and Pathway Enrichment Analysis

All choroid plexus tissues used for proteomic analysis shall be blood-free, as blood components may introduce nonspecific proteins and adversely affect the accuracy of downstream analyses. Thus, rats were transcardially perfused with ice-cold PBS under deep anesthetization by ketamine/xylazine (75:10 mg/kg, l mg/kg, ip.) as previously described. Blood-free choroid plexus tissues were carefully isolated from individual rat brains for protein extraction (n = 3 / age group). The following proteomic analyses were conducted based on the workflow developed by the Purdue Proteomics Facility (Hedrick et al. 2015; Mohallem and Aryal 2020). Collected tissues were washed once with PBS and then resuspended in 100 mM ammonium bicarbonate supplemented with protease inhibitors. The suspensions were then transferred to 2 mL vials with 1.4 mm ceramic beads and were lysed using a Precellys24 tissue homogenizer. The protein concentrations were determined by the BCA assay kit. Equivalent volumes to 50 μg of protein from each sample were precipitated with 4x volumes of cold acetone at −20 °C overnight. These samples were then centrifuged at 13,500 rpm for 10 minutes; after removing the acetone supernatant, pellets were dried by vacuum centrifugation for 15 minutes. The dried samples were then resuspended and solubilized in 60 μL urea (at 8 M)-DTT (at 10 mM) solution at 37 °C for 1 hour. Another 60 μL alkylation solution (97.5% Acetonitrile, 0.5% Triethyl phosphine, 2% Iodoethanol) was added and samples were incubated at 37 °C for another hour. These samples were then centrifuge-dried under a vacuum and resuspended in 20 μL of 0.05 μg/μL Lys-C/Trypsin in 25 mM ammonium bicarbonate.

The following digestion was conducted using a barocycler (50 °C, 60 cycles: 50 s at 20 kPSI and °C 10 s at 1 atmosphere). The yielded samples were further desalted by solid phase extraction with Min-spin C18 columns. Purified peptides were acquired by reversed-phase chromatography and 1 μg for each sample (measured by Nanodrop) were then injected to the liquid chromatography (LC)/mass spectrometry (MS)-MS system using the Dionex UltiMate 3000 RSLC nano System coupled to the Orbitrap Fusion Lumos Mass Spectrometer (Thermo Fisher Scientific, Waltham, MA). Search of raw MS/MS data was performed by MaxQuant software (v1.6.3.3) (Cox and Mann 2008) against Uniprot Rattus norvegicus database. Parameters for the search includes digestion by Lys-C/Trypsin, methionine oxidation, and alkylation of cysteine as fixed modification. The tolerance for peptide search was set at 10 ppm, and false discovery rate (FDR) of peptides and protein identification was set to 1%. Peptide-based label-free quantification (LFQ) algorithm was adopted for protein abundance measurement and proteins with at least 1 characteristic peptide and 2 MS/MS counts were used for further statistical analysis. Heat map, volcano plots, and Venn diagram depicting differentially expressed proteins among groups were generated by an interactive web platform developed by Monash University (Shah et al. 2019). Gene ontology tools were used to identify the enriched pathways by differentially expressed protein probesets between every two age groups (FDR set at 5%).

2.8. Statistics

All results are expressed as mean ± standard deviation (SD). Differences among multiple groups were statistically tested by one-way ANOVA with post hoc comparisons by the Dunnett’s test. Differences were considered statistically significant if p-values were equal or less than 0.05.

3. Results

3.1. Age-related Decline in Cu Clearance by the BCB

The in situ VC perfusion technique provides a tangible means to estimate the clearance of materials from the CSF by the choroid plexus (Al-Sarraf et al. 2000). Radioactive ⁶⁴Cu, along with ¹⁴C-sucrose, were continuously infused into the lateral ventricle, and the CSF effluent was subsequently collected from the cisterna magna to determine the eluted radioactivity. The loss of ⁶⁴Cu radioactivity in the cisternal CSF outflow, which was corrected by a space marker ¹⁴C-sucrose for interfering factors such as dilution and diffusion and compared with the ventricular inflow, reflected the fraction of ⁶⁴Cu removed from the CSF by the BCB efflux mechanism.

After a 70-min VC perfusion, the percentages of the recovered radioactivity in the CSF effluent were plotted against the perfusion time (Fig. 1B, C). The ¹⁴C-sucrose radioactivity in the CSF effluent plateaued at 30 min after the start of the perfusion in all age groups and remained at this level until the end of the experiment. At steady state, almost 100% of the ¹⁴C-sucrose was recovered in the CSF outflow, suggesting little ¹⁴C-sucrose was taken up by the choroid plexus. No statistical difference was determined among the tested groups for ¹⁴C-sucrose (Fig. 1D). Thus, the unchanged kinetics of the space marker, ¹⁴C-sucrose, indicated that the BCB integrity in all age groups remained intact.

By contrast, a significantly higher percentage of ⁶⁴Cu was recovered from the CSF effluent of old animals (48.8% ± 2.42%) as compared to young (39.6% ± 1.93%) and adult animals (39.1 % ± 0.97%), with a statistical significance (p < 0.05, n = 4) (Fig. 1D). After correction against ¹⁴C-sucrose, the ⁶⁴Cu recovery was found elevated by approximately 24% in the old animals’ CSF outflow samples as compared to the young and adult animals (Fig. 1D). These data suggest that a higher proportion of Cu failed to be cleared by the BCB efflux mechanism in the choroid plexus. We further calculated Cu clearance based on previous computational methods (Al-Sarraf et al. 2000). Our data demonstrated Cu clearance in choroid plexuses from old rats (29.3 ± 0.34 μl/min) were significantly reduced by 16% as compared to the young (35.1 ± 1.29 μl/min) and adult rats (34.7 ± 0.86 μl/min) (p < 0.05, n = 4) (Fig. 1E).

3.2. Altered mRNA Expression of Cu Transporters in Choroid Plexus of Different Ages

We conducted qPCR analyses to investigate the relative expressions of Ctr1, Dmt1, Atp7a, and Atp7b mRNA in the choroid plexus of rats at different ages. Our data revealed that Ctr1 mRNA expression in old choroid plexus was significantly upregulated, which was 1.6- and 2.9-fold higher than that of young and adult choroid plexus, respectively (p < 0.0001, Fig. 2A). Dmt1 expression in the choroid plexus from young rats was significantly downregulated, only 42.5 % and 36.5% of those in adult and old, respectively (p < 0.001, Fig. 2B). As for the two Cu exporters (i.e., Atp7a and Atp7b), age did not seem likely to alter the mRNA expression of Atp7a across all age groups; however, Atp7b expressions in the choroid plexus of adult and old animals were significantly lower than that in young animals, about 3.0- and 4.8- fold less, respectively (p < 0.0001, Fig. 2C, D). Overall, age affected the expression of mRNAs encoding Cu transporters, i.e., Ctr1, Dmt1, and Atp7b.

Figure 2. Altered mRNA expressions of Cu transporters in choroid plexus of different ages.

Relative mRNA expression levels of Ctr1 (A), Dmt1 (B), Atp7a (C), and Atp7b (D) in freshly isolated choroid plexus tissues of different ages were determined by qPCR. Data represent mean ± SD, n = 6; **: p < 0.01, ***: p < 0.001, and ****: p < 0.0001, as compared to the young; ###: p < 0.001, and ####: p < 0.0001, as compared to the adult. Gapdh was used as the reference gene to calculate relative changes across samples.

3.3. Subcellular Distribution of Cu Transport Proteins in Choroid Plexus of Different Ages

Following qPCR analyses, we used confocal imaging to investigate the subcellular distribution of Cu transporters in the choroid plexus. The choroid plexus images were acquired from microtome-cut brain slices covering the lateral ventricle posterior to the bregma. In young choroid plexus, CTR1 was expressed in the cytosol of CPECs, as indicated by co-staining of CTR1 and TTR; however, discontinuity of CTR1 expression on the apical side was also observed. Noticeably, CTR1 also expressed in the stromal space between CPECs and choroidal capillary (with negative TTR staining) showing a speckled distribution (Fig. 3A upper panel). In contrast to young choroid plexus, the adult showed a continuous expression of CTR1 along the apical region but much less in the basolateral region. Moreover, CTR1 was also clearly seen in choroidal endothelial cells in the stromal space, suggesting a role of CTR1 in Cu transport between the blood and CSF (Fig. 3A middle panel). The observation of CTR1 in adult choroid plexus is in agreement with previous literature (Haywood and Vaillant 2014). In old choroid plexus, CTR1 expression in CPECs appeared to be weaker as compared to the young and old choroid plexus, although the existence on the apical side remained evident (Fig. 3A lower panel). The weak expression of CTR1 in the old was further verified by WB analyses. The WB data revealed a decreased expression of CTR1 in old choroid plexus by 29.4% (p = 0.068) and 32.1% (p < 0.05) compared to the young and the adult choroid plexus, respectively (Fig. 3E and 3F). A declined CTR1 expression in the old plexus tissue appeared to suggest age-related CTR1 degeneration.

Figure 3. Expression of Cu transporting proteins in the choroid plexus of different ages.

IHC data revealed the expression of CTR1 (A), DMT1 (B), ATP7A (C), and ATP7B (D) in choroid plexus tissues of young, adult, and old rats. Scale bar = 50 μm. Representative WB images were provided in (E). Semi-quantitative analyses of Cu transporting protein expression levels were in (F), (G), (H), and (I). Data represent mean ± SD, n = 3; #: p < 0.05, as compared to the adult.

DMT1, a weaker Cu uptake protein (Zheng et al. 2012), distributed mainly to the CPEC cytoplasm in the young, showing no disparity between the apical and basolateral aspect. Interestingly, little DMT1 expression was seen in the stroma. For the adult, DMT1 was found more on the apical than the basolateral side, with significant signals seen in the stroma, presumably in the choroidal endothelia. In contrast, DMT1 in the old choroid plexus was more concentrated on the apical side (Fig. 3B). The WB data, however, did not reveal any significant difference in DMT1 expression in the choroid plexus among the tested age groups (Fig. 3E and 3G).

ATP7A and ATP7B are Cu subcellular trafficking proteins that were both normally expressed in the perinuclear areas of adult choroidal epithelial cells (Fu et al. 2014). Our data by IHC staining were consistent with past observations across all age groups with the following characteristics. First, ATP7A expression in young choroid plexus was evenly distributed in the CPEC cytoplasm, while in adult and old ATP7A was more concentrated on both apical and basolateral side (Fig. 3C). Second, ATP7B was also expressed in the stromal space, yet its intensity seemed to be lower than other Cu transporting proteins (Fig. 3D). Third, although no significant changes were found by the WB, ATP7B expression in old choroid plexus was 24.6% and 17.9% higher than that in the young and the adult, respectively (Fig. 3E, 3H, 3I). Hence, these data suggested a distinct age-specific expression of Cu transporters in the choroid plexus.

3.4. Changes in mRNA Expression of Tight Junction Genes in Choroid Plexus of Different Ages

We used qPCR to investigate the age effect on genes encoding tight junction (TJ) proteins in the choroid plexus. Zo-1 mRNA was highly expressed in adult choroid plexus, which was 5.6- and 3.7- fold higher than those of young and old choroid plexuses, respectively (p < 0.0001, Fig. 4A). A similar trend was also observed in Claudin-12 mRNA expression: its relative expression in adult choroid plexus was 1.8- and 2.1- fold higher than that of young and old choroid plexuses, respectively (p < 0.01, Fig. 4D). The mRNA expression of Occludin and Claudin-5 shared a similar pattern, where these two mRNAs were significantly highly expressed in young choroid plexus compared to that of other age groups (Fig. 4C, D). In addition, Connexin-43 was also found to be highly expressed in the young choroid plexus: its relative expression in young choroid plexus was 2.6- and 2.8- fold higher than those in adult and old choroid plexuses, respectively (p < 0.01, Fig. 4F). Zo-2 mRNA expression was not altered by age across different age groups (Fig. 4B).

Figure 4. Expressions of mRNAs encoding tight junction proteins in the choroid plexus of different ages.

Relative mRNA expression levels of Zo-1 (A), Zo-2 (B), Occludin (C), Claudin-12 (D), Claudin-5 (E), and Connexin 43 (F) in young, adult, and old animals. Data represent mean ± SD, n = 6; *: p < 0.05, **: p < 0.01, ***: p < 0.001, and ****: p < 0.0001, as compared to young; ##: p < 0.01 and ###: p < 0.001, as compared to adult. Gapdh was used as the reference gene to calculate relative changes across samples.

3.5. Proteomic Profiling of Choroid Plexus of Different Ages

We found that 81 out of 3414 identified proteins (2.37%) were differentially expressed across the three age groups. The hierarchical clustering of differentially expressed proteins exhibited a distinct proteome pattern in young choroid plexus compared to the other two groups, while the proteomes between adult and old choroid plexus only showed slight differences on a few clusters (Fig. 5A). The heatmap with identified protein annotations is also provided in the Supplemental Material 1 in the svg format. Further “volcano” plots depicting the changes between every two age groups validated the differences observed in the heatmap (Fig. 5B-D). For example, a total of 67 proteins expressed differentially between the young and old choroid plexus (Fig. 5B) and 36 between the young and adult choroid plexus (Fig. 5C), while only 8 proteins differentially expressed between the adult and old choroid plexus (Fig. 5D). These findings are summarized in the Venn diagram (Fig. 5E), highlighting the novel aspects of proteomes by the young choroid plexus. The majority of the differentially expressed proteins (73 out of 81) were identified in young choroid plexus, while old choroid plexus expressed over half of (45 out of 81) differentially expressed proteins. Interestingly, however, no protein specific to the adult-vs-old or young-vs-adult-vs-old was found. These data suggest proteome differences between young-vs-old and between young-vs-adult may underlie the functional differences between these groups. The complete proteomic dataset and statistical analyses can be found in Supplemental Material 2 and 3, respectively.

Figure 5. Proteomic profiling of young, adult, and old rat choroid plexus tissues.

(A). Heatmap of 81 proteins differentially expressed among young, adult, and old choroid plexus. (B-D). Volcano plots depicting group-wise differences of differentially expressed proteins. Proteins with log2 fold-changes beyond 1 or below −1 with adjusted p values lower than 0.05 were extracted to show the age-related expression changes. (E). Venn diagram highlighting the overlaps between the differentially expressed proteins identified in the volcano plots as shown in (B), (C), and (D).

3.6. Pathway Enrichment Analysis

The proteomic data, was further analyzed via GENE ONTOLOGY (GO) to investigate age-associated differences in protein expression at the pathway-level. The GO analyses were conducted on three major categories: cellular component, molecular function and biological process. The young-old comparison determined pathway-level differences in all three categories with the following characteristics. First, among eleven cellular component-related pathways identified, there existed distinct age-related differences in collagen-associated extracellular matrix, especially on the basolateral side of choroid plexus (Fig. 6A). Second, six pathways involving molecular functions differed significantly between the young and old choroid plexuses (Fig. 6B); these pathways were also pertinent to collagen-associated extracellular matrix alterations. Noticeably, the biomacromolecule binding capability in the young choroid plexus was significantly different from the old. Third, three pathways featuring biological processes were also significantly different between the young and old choroid plexuses (Fig. 6C). Alterations in the choroid plexus extracellular matrix were once again prominent, with one development-related pathway also identified. Finally, the comparison made between the young and adult choroid plexuses revealed a difference only in cellular component-related pathways (Fig. 6D), where the variations were mainly in polarized growth, extracellular space, mitochondrial function, and possibly the regulation on other brain regions. Overall, our proteomic analysis suggested an age-associated difference in the protein expression pattern in the choroid plexus, which may contribute to Cu transport regulation in the BCB.

Figure 6. Pathway enrichment analysis for the Young vs. Old or Young vs. Adult comparisons.

The probeset list was uploaded to the GO tool to investigate the significantly different pathways (false discovery rate, i.e., FDR, < 0.05). Resultant pathways were ranked by the −log(p-value). Pathways were highlighted for group comparison made between young and old choroid plexus in categories of cellular component (A), molecular function (B), and biological process (C, D). The pathways were highlighted between young and adult choroid plexus on cellular component.

To verify an age-related disruption in extracellular matrix in the choroid plexus as suggested by the proteomic analysis, we investigated alterations in collagen deposition dependent upon age. Collagen IV, a frequently researched collagen type in choroid plexus (González-Marrero et al. 2015; Urabe et al. 2002; Wei et al. 2000), was selected as a marker for this purpose. Our confocal data showed collagen IV was present not only on the basolateral side of the CPECs but also on the choroidal endothelial layer, with the latter having a greater deposition (Fig. 7A). The old choroid plexus showed significantly more collagen IV on both the CPEC basolateral side and the choroidal endothelia than did young and adult choroid plexuses (Fig. 7A). Quantitative analyses revealed that the thickness of collagen deposition at the basolateral CPECs in old animals (1.30 ± 0.08 μm) were significantly higher than that in the young (0.80 ± 0.02 μm) and the adult (0.84 ± 0.02 μm) by 62.5% and 54.8%, respectively (n = 3, p < 0.001) (Fig. 7B). In addition, the thickness of collagen deposition in the old choroidal endothelium (2.60 ± 0.20 μm) was also significantly greater than that in the young (1.30 ± 0.08 μm) and the adult (1.40 ± 0.11 μm) by 100.0% and 85.7%, respectively (n = 3, p < 0.001 compared to the young, p < 0.0001 compared to the adult) (Fig. 7C). Noticeably, the collagen deposition thickness in choroidal endothelia nearly doubled that in the basolateral CPECs in every age group. Finally, co-localization of collagen (green) and TTR (red), which yields the yellowish fluorescence, was observed only in CPECs of old choroid plexus, but not in young and adult (Fig. 7A lower panel), suggesting that an abnormal collagen deposit occurred inside of old CPECs. Overall, our IHC data on collagen IV from brain slices were in a good agreement with the proteomic data reported in Fig. 6A-D.

Figure 7. Collagen IV staining in the choroid plexus.

Expression of Collagen IV in young, adult and old choroid plexus tissues was investigated using IHC staining (A). The thickness of collagen deposition in basolateral CPECs (B) and choroidal endothelia (C) were quantitatively analyzed. Data represent mean ± SD, n = 3; ****: p < 0.0001, as compared to the young; ###: p < 0.001, and ####: p < 0.0001 as compared to the adult.

4. Discussion

Our current study provides the clear evidence that: (1) Cu clearance at the BCB of old rats declined significantly as compared to the young and adult animals; (2) age-related expression pattern of Cu transporters and junctional proteins in the choroid plexus exist; and (3) the choroid plexus possesses different proteomes due to age, particularly in pathways pertinent to extracellular matrix.

Age has been suggested as an independent factor affecting brain Cu levels. Multiple lines of evidence have revealed age-dependent brain accumulation of Cu (Ashraf et al. 2019; Fu et al. 2015; Sathyanarayana Rao and Rao 2010). Although essential to brain health, Cu excess has been associated with neurodegenerative diseases, especially the Alzheimer’s disease (AD) (Bagheri et al. 2018). Cu induces the formation of β-sheet and α-helix structure of amyloid peptides, creates an acidic environment required by fibril formation, and induces free radicals by Fenton-like reactions. Overall, these Cu-induced effects in these overload scenarios, are capable of promoting the aggregation-prone kinetics of β-amyloid, which further leads to AD pathogenesis (Bagheri et al. 2018). In addition, Cu overload has been reported in idiopathic Parkinson’s disease (PD), with higher Cu concentration detected in CSF samples of PD patients than those of control subjects (Pall et al. 1987). Epidemiological evidence in occupational populations has also established the association between Cu overexposure and PD (Gorell et al. 1999). Given the existing evidence showing higher brain Cu burden and increased incidence of neurodegenerative diseases in older individuals, it is imperative to explore the mechanisms by which Cu dyshomeostasis takes place in an age-dependent manner.

Two brain barrier systems, i.e., the blood-brain barrier (BBB) and BCB, control Cu transport between the cerebral milieu and peripheral blood circulation. Net brain Cu levels are ultimately determined by Cu influx into and Cu efflux out of the brain. Research by our group has highlighted the respective duties of Cu influx by the BBB and Cu efflux by the BCB (Fu et al. 2014). A longitudinal study has shown a trend of increased brain Cu uptake in older brains by ⁶⁴CuCl₂-positron emission tomography/computed tomography (PET/CT) imaging (Peng et al. 2018); however, the contribution of BCB were not investigated. Compared to the BBB, the choroid plexus exhibits a significantly higher capability for Cu transport (Choi and Zheng 2009). The current study utilizing VC perfusion to assess Cu efflux at the BCB showed that the BCB’s Cu clearance rate was significantly decreased in old rats as compared to young and adult animals, which could contribute, at least in part, to age-dependent Cu increases in the brain. Noticeably, the old rats used in this study were about 18 months old, an age that is still far from the expected rat life span (3 years) (Pallav Sengupta 2013). Hence, we postulate that the Cu clearance at the BCB would further decline in the elderly, leading to an even greater Cu dyshomeostasis in their brains. This hypothesis deserves further experimental investigation.

Cu transporters in the choroid plexus determine the rate and direction of Cu fluxes between the CSF and blood, and their expression in the choroid plexus could be affected by age. CTR1 and DMT1 are two major Cu uptake proteins that acquire Cu from the extracellular matrices into the choroidal epithelial cells. Our IHC and Western blot data determined CTR1 expression in old animals was significantly decreased compared to the young and adult animals, suggesting diminished Cu uptake from the CSF. DMT1, another Cu importer with less Cu uptake capacity compared to CTR1 (Zheng et al. 2012), showed a perinuclear distribution in the young choroid plexus. The adult and old choroid plexus, however, exhibited DMT1 distribution more concentrated on the apical side. Therefore, it appeared likely that DMT1, similar to CTR1, also favored Cu removal from CSF to blood in old animals, although the aging process does not apparently alter DMT expression in the choroid plexus. ATP7A and ATP7B, by regulating intracellular Cu trafficking, function as Cu exporters in response to cellular Cu overload (Fu et al. 2014). In the young choroid plexus, both were expressed in cytoplasm, indicating a balanced Cu level between CSF and blood. However, both were translocated to the apical sides in the adult and old choroid plexus. Interestingly, Western blot assessments did not reveal any significant changes in the expression of ATP7A or ATP7B in the old choroid plexus. But the non-significant increased expression of ATP7B, as revealed by WB, may still in part contribute to the decreased Cu efflux by exporting Cu to the CSF. Overall, it is reasonable to postulate that decreased CTR1 expression in the old choroid plexus is a primary factor contributing to diminished clearance of Cu from the CSF in old rats.

As a barrier between the blood and CSF, tight junctions between adjacent choroidal epithelial cells are of importance in preventing the entrance of blood-borne substances into the CSF. Our ¹⁴C-sucrose data from the VC perfusion experiment suggested intact BCB integrity in all groups. It was therefore surprising that there existed age-related alterations in mRNA expression of several critical junctional proteins. Relative expressions of mRNAs encoding major tight junctional proteins were significantly different across all age groups except for Zo-2. How the age may affect the expression of tight junction proteins in the BCB remains unknown. It is possible that the epigenetic mechanism reported in other epithelial systems (Khan and Asif 2015; Morini et al. 2018) may modulate the junctional protein expression and thus help regulate the BCB permeability. Given the decreased expressions of Zo-1 and Claudin-12, two crucial tight junction genes, we propose BCB permeability may decline as age advances. A structurally impaired the BCB barrier in the elderly may compromise the barrier’s function, reducing clearance of substances. This may contribute to Cu dyshomeostasis in the elderly.

Our proteomic data demonstrated age-specific proteomes in the choroid plexus, although the differences existed mainly between the young and old age groups. Among various pathways revealed by the pathway enrichment analysis, the extracellular matrix- and collagen-related proteomes exhibited distinctly different patterns between the young and old groups. The post-validation by collagen IV staining was consistent with the major hints from proteomics. A significant collagen thickening was observed in old choroid plexus, not only on the basolateral CPECs but also around the choroidal endothelium. This observation is in a good agreement with previous report in literature (González-Marrero et al. 2015), who used the 16-month-old triple transgenic AD mouse model and observed a collagen-related thickening of the epithelial basal membrane in the choroid plexus and a greater collagen-IV deposition around the choroidal capillaries. These authors postulate that changes may curtail the solute exchange, resulting in an insufficient clearance of Aβ from the CSF to blood. While it is not the focus of this study, we suspect that the extracellular matrix in the old choroid plexus, as revealed by our proteomic study and collagen staining by IHC, may impact Cu movement from choroidal epithelial cells to the blood. This hypothesis, however, needs further experimental testing.

As a barrier between two distinct body fluid compartments (i.e., blood and CSF), the choroidal epithelial cells possess a unique polarity which allows for transporting proteins to distribute either towards the apical microvilli facing the CSF or close to the basolateral membrane facing the blood (Christensen et al. 2018; Yonemura 2014). Such a cellular polarity is presumably maintained by the extracellular matrix and intracellular protein sorting system (Tharp and Weaver 2018) and is deemed important to the distribution of Cu transporters. Indeed, the data from our own previous studies have showed that the cupper transport protein-1 (CTR1) normally distributes between the nuclei and apical membrane in the choroid plexus; in vivo chronic Mn exposure in rats causes the relocation of CTR1 to the apical brush board of choroidal epithelia (Zheng and Monnot 2012). In another study, we have established that both ATP7A and ATP7B, two prominent Cu-expel proteins, were mainly distributed in the perinuclear region of the choroidal epithelia; the presence of excessive Cu induced by manganese (Mn) overexposure promotes the subcellular trafficking of ATP7A toward the apical microvilli, yet it translocates ATP7B in the opposite direction toward the basolateral membrane (Fu et al. 2014). Literature by Jain et al. (2015) further demonstrates that ATP7B polarized sorting does exist and functions to maintain the neuronal Cu homeostasis.

With regards to the fact that the old choroid plexus had a declined Cu efflux capability when compared to the adult, it was surprising that few pathways were different between these two age groups. Recent data from literature, however, suggest that factors affecting choroid plexus metabolism may influence Cu transporters; for example, a senescence-induced autophagy over-activation can cause abnormal expressions of Ctr1 and Atp7a, leading to intracellular Cu overload (Masaldan et al. 2018). Thus, a longitudinal study to elucidate the interactions between autophagy and intracellular sorting of Cu transport proteins in choroid plexus is necessary for better understanding of age-dependent changes.

5. Conclusion

In summary, our findings demonstrate a decreased Cu clearance at the BCB in old rats. Specifically, there is an altered, age-dependent expression pattern with regards to Cu transporters in the choroid plexus. Proteomic analyses of choroid plexus tissues illustrate a distinctly different proteomes in the old choroid plexus, particularly involving the extracellular matrix-related pathways. These observations provide the firsthand evidence to support an age-dependent Cu dyshomeostasis in the CSF, which is regulated by yet-to-be- demonstrated mechanisms at the blood-CSF barrier in the choroid plexus. The major findings in this study are summarized and graphically illustrated in Fig. 8.

Figure 8. Graphical Illustration of Age-dependent Decline of Copper Clearance at the Blood-Cerebrospinal Fluid Barrier.

Under normal condition in young and adult choroid plexus, Cu ions are removed by CTR1 situated on the apical surface of the choroidal epithelia; the clearance from the CSF is facilitated by other metal transporters such as ATP7A and APT7B. Aging diminishes the expression of CTR1, slightly upregulates the apical ATP7B expression, increases the deposition of collagen, and alters proteomes, particularly in the pathways involving the extracellular matrix. These age-dependent changes in choroid plexus ultimately contribute to Cu dyshomeostasis in the CSF.

Highlights.

• Cu clearance from CSF by the choroid plexus was diminished in aged animals.

• Cu transporters in choroid plexus exhibited an age-specific expression pattern.

• Choroid plexus has age-specific proteome revealed by proteomic analyses.

• Disturbed extracellular matrix in old choroid plexus may undermine BCB Cu efflux.

Abbreviation:

Cu

copper

CSF

cerebrospinal fluid

BCB

blood-CSF barrier

VC perfusion

ventriculo-cisternal perfusion

CTR1

high affinity copper uptake protein 1

DMT1

divalent metal transporter 1

ATP7A

ATPase copper transporting alpha

ATP7B

ATPase copper transporting beta