The extent of damage to the blood-brain barrier in the hypercholesterolemic LDLR−/−/Apo E−/− double knockout mice depends on the animal's age, duration of pathology and brain area

One of the effects of hypercholesterolemia (Hch) exerted on the central nervous system (CNS) is damage to the blood-brain barrier (BBB). Increased permeability of BBB results from structural changes in the vascular wall, loss of the tight junctions and barrier function, as well as alterations in the concentration of proteins located in the layers of the vascular wall. These changes occur in the course of metabolic and neurodegenerative diseases. The important role in the course of these processes is attributed to agrin, matrix metalloproteinase-9, and aquaporin-4. In this study, we aimed to determine: 1) the extent of Hch-induced damage to the BBB during maturation, and 2) the distribution of the above-mentioned markers in the vascular wall. Immunohistochemical staining and confocal microscopy were used for vascular wall protein assessment. The size of BBB damage was studied based on perivascular leakage of fluorescently labeled dextran. Three- and twelve-month-old male LDLR-/-/Apo E-/- double knockout mice (EX) developing Hch were used in the study. Age-matched male wild-type (WT) C57BL/6 mice were used as a control group. Differences in the concentration of studied markers coexisted with BBB disintegration, especially in younger mice. A relationship between the maturation of the vascular system and reduction of the BBB damage was also observed. We conclude that the extent of BBB permeability depends on animal age, duration of Hch, and brain region. These may explain different susceptibility of various brain areas to Hch, and different presentation of this pathology depending on age and its duration.


Introduction
Disorders of lipid metabolism, including hypercholesterolemia (Hch), are at the base of civilization diseases, such as atherosclerosis, arterial hypertension, stroke, and cardiovascular diseases. Together, they are the second major cause of death, and the third leading cause of morbidity and disability in the population of developed countries worldwide (Hankey, 2017;Ortiz-Prado et al., 2021). The latest reports also indicate a close relationship between lipid disorders and the development of neurodegenerative diseases, such as Alzheimer's disease (AD), Parkinson's disease, and Huntington's disease (Huang et al., 2019;Jin et al., 2019;Kivipelto et al., 2002b;Scheltens et al., 2021;Wu et al., 2022). In addition, familial hypercholesterolemia (FH) has been associated with mutations in the low-density lipoprotein receptor gene, which promotes elevated plasma LDL levels (Defesche et al., 2017). High levels of LDL by disrupting neurogenesis in the hippocampus may be responsible for cognitive impairment (Engel et al., 2019).
Disturbances in the brain cholesterol metabolism are associated with changes in the concentration of its intermediate metabolites and the overproduction of oxysterols (Gosselet et al., 2014;Hughes et al., 2013), which may induce dysfunction of all components of the neurovascular unit (NVU) (Hawkes et al., 2015;Kowiański et al., 2013;Zlokovic, 2008). Therefore, among the most important consequences of Hch are damage to the cerebral arterial walls and increased permeability of the blood-brain barrier (BBB). Vascular wall components such as endothelial cells, basement membrane, muscular layer, astrocytes perivascular endfeet, and pericytes contain specific markers whose expression is altered in the cerebrovascular and neurodegenerative diseases (Claudio et al., 1995;Farkas and de Jong, 2000). Among the important structural markers of the vascular wall are: agrin (Agr), matrix metallopeptidase-9 (MMP-9), and aquaporin-4 (AQP4) (Fig. 1). Their role is to maintain the NVU structure and, in particular, the integrity and function of BBB (Zlokovic, 2008).
Agr is a heparan sulphate proteoglycan, whose secreted form is expressed in the basal lamina of cerebral vessels, whereas the transmembrane form is highly expressed in the brain tissue (Barber and Lieth, 1997;Kröger, 1996). In physiological conditions, the secreted form of Agr is involved in the formation and maintenance of BBB (Liebner et al., 2011;Smith and Hilgenberg, 2002). Agr shapes vascular endothelial junctions by regulating the concentration of junction proteins (Steiner et al., 2014). Its concentration in the microvascular basal lamina affects the astrocyte polarity, which is important for BBB integrity (Liebner et al., 2011). Agr is involved in angiogenesis, and its accumulation in the basal lamina correlates with BBB maturation (Steiner et al., 2014). Under physiological conditions, the transmembrane form of Agr regulates the excitatory synaptic activity and synaptic plasticity (Li et al., 1997;O'Connor et al., 1994). It contributes to the synaptic formation in the mature hippocampus (Porten et al., 2010). Decrease in Agr level can lead to cognitive deficits and impairment of spatial memory Skucas et al., 2011;Zhang et al., 2013).
AQP4 is the most abundant water channel in the perivascular membranes of the astrocytic endfeet (Mader and Brimberg, 2019a). It is also highly expressed in the ependymal cells of the choroid plexus and the reactive astrocytes (Gundersen et al., 2014). Under physiological conditions, brain water dynamics is precisely regulated by the perivascular AQP4 pool. Altered AQP4 expression impairs water transport leading to the development of cytotoxic edema and potentially also neurological disorders (De Oliveira et al., 2014;Mader and Brimberg, 2019b;Wolburg et al., 2012). de Oliveira et al. reported the increased AQP4 immunoreactivity in the hippocampus of the LDLR − /− mice, which increased their susceptibility to BBB disruption (De Oliveira et al., 2014). Moreover, the LDLR − /− mice exposed to a high-cholesterol diet were more susceptible to cognitive impairment.
Another marker of the vascular wall is MMP-9, which represents the superfamily of proteolytic enzymes. It is expressed in neurons, microglia, oligodendrocytes, endothelial cells, and vascular smooth muscle cells, in several brain regions (Aujla and Huntley, 2014;Bednarek et al., 2009;Oliveira-Silva et al., 2007;Vaillant et al., 1999). MMP-9 degrades not only the extracellular matrix, but also cell adhesion molecules, cell surface receptors, and other proteases (Reinhard et al., 2015). Under physiological conditions, MMP-9 is involved in the maintenance of NVU integrity and degradation of amyloid β (Aβ) fibrils, contributing to the clearance of amyloid plaques (Yin et al., 2006). Excessive activation of MMP-9 results in impairment of Aβ clearance, which can initiate neurodegeneration. In Hch, MMP-9 plays an important role related to induction of the cerebral vessel remodeling (Czuba et al., 2017;Deng et al., 2014). The increased expression of MMP-9 correlates with the increased cerebral vascular tortuosity, decreased inner diameters of the middle cerebral artery, and increased microvessel density in mice (Deng et al., 2014).
Despite growing evidence showing that changes in the expression of the vascular wall markers in the course of Hch are at the base of BBB Fig. 1. Schematic presentation of the cerebral vascular wall section. Localization of the proteins responsible for maintaining the functional blood-brain barrier is presented. Agr is expressed in the basal lamina of cerebral vessels, AQP4 is in the perivascular membranes of the astrocytic endfeet, the ependymal cells of the choroid plexus and the cytoplasm of reactive astrocytes, and MMP-9 is expressed in neurons, microglia, oligodendrocytes, endothelial cells, and vascular smooth muscle cells. α-Dystroglycan is bound to β-dystroglycan to form the dystrophin-associated protein complex (DAPC), which is an integral protein in the cell membrane. α-Dystroglycan binds to the extracellular components of the basal lamina such as laminin and agrin using β-dystroglycan. The cytoplasmic complex interacts with proteins that are directly linked to the AQP4 channel in the plasma membrane, anchoring AQP4 to astrocytic endfoot membrane domains. MMP-9, present in the brain microvasculature, is a linker between astrocyte endfeet and parenchymal basement membrane molecules. Ocludin and claudin are responsible for maintaining integrity of tight junctions between endothelial cels. E. Czuba-Pakuła et al. damage (de Bem et al., 2021;, still several issues should be addressed. It remains an open question whether the Hchinduced BBB changes are equally intense in all brain areas and whether their intensity is related to the length of pathology. It is not known whether changes in the expression of the above-mentioned markers are related to the rate of BBB leakage and cortical or subcortical localization of arterial branches. Finally, it is not known if the rate of Hch-induced BBB leakage depends on age. Hence, in this study, we aimed to determine the effect of Hch on the vascular wall and BBB permeability in various brain areas during its maturation, based on: 1) the assessment of the Hch-induced damage to the BBB in selected cortical and subcortical brain areas, and 2) the assessment of the distribution of Agr, MMP-9, and AQP4 in the vascular wall of hypercholesterolemic versus control mice.

Animals
Ten male LDLR − /− /Apo E − /− double knockout mice (EX) at the age of 3 months and ten mice at the age of 12 months, developing Hch and atherosclerosis (Getz and Reardon, 2016), as well as 20 male agematched wild-type (WT) C57BL/6 mice were used in this study. All animals were obtained from the Department of Biochemistry and Tri-City Academic Laboratory Animal Centre -Research and Services Centre at the Medical University of Gdansk. The mice were housed in individually ventilated cages, not exceeding 5 individuals per cage, with an area of 432cm 2 (22.5 ± 0.5 • C, 40 ± 5 % humidity) in 12-h alternating day/night cycles with unlimited access to clean water and standard chow. Twenty of the animals were used for BBB permeability studies and twenty for immunohistochemical (IHC) studies. The project was accepted by the Local Ethical Committee of the Medical University of Gdansk (WHiBZ/lke.003/29/18). All animal handling procedures and experimental protocols were performed under the provisions of the EU Council Directive 2010/63/EU for animal experiments, with the preservation of humanitarian care and the use of laboratory animals to minimize animals' pain and discomfort and to reduce the number of experimental subjects. The study is reported according to the ARRIVE guidelines (Percie du Sert et al., 2020).

BBB permeability assays with FD40 and tissue preparation
The BBB permeability was determined using 40-kDa fluorescein isothiocyanate-dextran (FD40; Sigma, UK) injected into a tail vein in the volume of 200 μl in 50 mg/ml concentration in NaCl 0.9 % (pH 7.4). The syringe was covered providing low light conditions to minimize the quenching of fluorescence of the FITC-dextran solution (Liu et al., 2018;Morita and Miyata, 2012). 60 min after the FD40 injection the animals were deeply anesthetized with a lethal dose of ketamine (100 mg/kg) and xylazine (10 mg/kg), administered intraperitoneally. Then, they were perfused transcardially with 150 ml solution of NaCl 0.9 % (pH 7.4) and with 150 ml of 4 % paraformaldehyde in phosphate-buffered saline (PBS; Sigma, UK; 0.1 M; pH 7.4) at room temperature. After removing from the skull, brains were fixated by immersion in a solution of 4 % paraformaldehyde for 2 h. Subsequently, they were subjected to a cryoprotection by immersion in 15 % and 30 % sucrose solutions in 0.1 M PBS (Sigma; pH 7.4; 4 • C) until they sank. After freezing at − 20 • C, the brains were cut into 40 μm thick sections on a freezing cryostat (Thermo Scientific Cryostat Microm HM 525, Germany) and stored in an antifreeze solution (− 80 • C) for further analysis. The sections were mounted on microscope slides, dried, and coated with either Kaiser's gelatin (Sigma, UK).

IHC staining protocol for Agr, MMP-9, and AQP4
The procedure for handling the animals from the moment of anesthesia to obtaining sections from the freezing cryostat was the same as described in Section 2.2. The free-floating sections were washed three times in PBS and placed in a blocking solution containing 5 % goat normal serum/0.3 % Triton X-100/PBS for 30 min at room temperature to block a non-specific binding. Subsequently, the sections were incubated with primary antibodies: rabbit polyclonal against Agr (Santa Cruz Biotechnology, CA; dilution 1:100) with mouse monoclonal against MMP-9 (Santa Cruz Biotechnology, CA; 1:300), and rabbit polyclonal against AQP4 (Santa Cruz Biotechnology, CA; 1:300) with MMP-9, respectively, for 24 h at 4 • C. The sections were washed again three times in PBS and incubated with the following secondary antibodies: goat anti-rabbit conjugated to Cy3 (Invitrogen, USA; 1:500) and goat anti-mouse conjugated to Alexa Fluor 488 (Thermo Fisher Scientific Inc., USA; 1:300) for 2.5 h at room temperature. Trials using a solution of Protein Block were also performed to exclude non-specific binding (BioGenex, CA). Finally, slides were washed in PBS, mounted on microscope slides, dried, and coated with Kaiser's gelatin (Sigma, UK).

Verification of staining specificity
The specificity of the IHC staining was verified by means of the omission tests. Briefly, the previously described IHC procedure was done, except that the primary or secondary antibody was omitted. These tests excluded non-specific stainings.

Microscopy imaging
The positively-stained structures were visualized and documented with the confocal laser scanning microscopy system LSM 880 (Zeiss, Germany) mounted on a microscope AxioImager.Z2 (Zeiss, Germany). For each animal, 10-15 coronal sections through the telencephalic region were selected under 10× magnification. The assessment of the BBB permeability assays with FD40 and IHC stained sections was done in the prefrontal (PFCx) and motor cortex (MCx), hippocampus (HIP), and striatum (STR). The selection of brain regions and structures was made based on the Mouse Brain in Stereotaxic Coordinates (Paxinos and Franklin, 2019). Images were obtained with a 20 × objective lens with a zoom of 1.0 or 2.8 times. Verification of the IF-stained sections in the non-corresponding channels eliminated the suspicion of the spectral bleed-through. The morphometric assessment was carried out on the obtained images using the image analysis program Zen 2.3 (Blue Edition; Zeiss).
In order to assess the size of the BBB leakage, the following quantitative parameters were studied: 1) the internal diameter of the vessels, 2) the mean diameter of the extravasation zone, 3) the maximal and 4) average fluorescence intensity. All measurements were made using the image analysis program Zen 2.3 (Blue Edition; Zeiss) -the results were obtained in micrometers [μm]. Additionally, in the case of the FD40 marker, also the appropriate intensity of fluorescence (arbitrary units) was measured.
Analysis of sections stained IHC against Agr, MMP-9, and AQP4 concerned the following quantitative parameters of the cerebral vessels: 1) the external diameter of the vessels, 2) internal diameter of the vessels, and 3) the mean thickness of the vascular wall. All measurements were made using the image analysis program Zen 2.3 (Blue Edition; Zeiss). The results were obtained in micrometers [μm].
For each of the examined brain structures stained with FD40, and IHC against Agr, MMP-9, and AQP4, measurements were made on 50 blood vessels (10 from each studied mice) in each of four studied structures (i.e. PFx, MCx, HIP and STR) and groups (3-and 12-month-old EX, and 3-and 12-month-old WT).

Statistical analysis
All analyses were performed using the STATISTICA version 13.3 (2020, StatSoft. Inc., USA). To verify the normality of data distribution, E. Czuba-Pakuła et al. the Shapiro-Wilk, Kołomogorow-Smirnow, and Jarque-Bera tests were used. The Leven (Brown-Forsythe) test was used to test the hypothesis of equal variances. Comparisons of mean values between groups were tested by Student's t-test or Mann-Whitney U test, as appropriate. The significance of differences between the same variable in different structures, in the absence of a normal distribution of the variable, was tested with the Kruskal-Wallis test. Statistical significance was assumed at p < 0.05. Data are presented as mean ± standard deviation (SD).

Changes in BBB permeability and size of the extravasation zone differ among hypercholesterolemic and wild-type mice of different age and cortical or subcortical localization of the studied area
The results showed an extravasation zone of fluorescently labeled FD40 in EX 3-month-and 12-month-old mice in PFCx, MCx, HIP, and STR (Fig. 2). Small leakage was also detected in 3-month-old mice from the control group. Whereas in EX mice of both studied age groups, FD40 was completely extravasated, in control mice it was also observed within the cerebral vessels. The assessment of the extravasation zone was based on four parameters: 1) the internal diameter of the leaky vessel, 2) the mean diameter of the extravasation zone, 3) the maximal fluorescence intensity, and 4) the average fluorescence intensity.

The internal diameter of the hypercholesterolemic leaky vessels differs in cortical and subcortical structures
In 3-month-old EX mice, BBB damage and perivascular leakage were found in all studied structures. In cortical structures (e.g. PFCx, MCx, and HIP) they were observed in vessels of larger internal diameter, compared with the vessels of animals from the control group ( Fig. 3A-C). However, there was no difference in the internal diameter of vessels with BBB damage, and perivascular leakage between EX and WT mice in STR (p = 0.6404, Fig. 3D).

The mean diameter of the extravasation zone is larger in younger mice with hypercholesterolemia
The mean diameter of the extravasation zone was larger in all the studied brain areas in 3-month-old EX mice compared with the control group (p < 0.0001; Fig. 4). On the contrary, in 12-month-old mice, the value of this parameter was greater only in HIP and STR compared with the control (p < 0.00001 and p < 0.01, respectively; Fig. 4C-D). While comparing two EX age groups, the mean diameter of the extravasation zone was larger in all studied structures of younger mice. Comparison of the mean diameter of the extravasation zone between mice from the WT groups at 3-month-old and 12-month-old showed no significant changes.

The maximal and average fluorescence intensity are higher in younger mice with hypercholesterolemia
The maximal intensity of fluorescence within the extravasation zone Fig. 2. Representative confocal scanning microscope images of cerebral blood vessel staining to 40 kDa FITCdextran (FD40) in 3-month-old and 12-month-old mice with hypercholesterolemia (EX) and control group (WT) in the prefrontal cortex (PFCx), the motor cortex (MCx), hippocampus (HIP) and striatum (STR). The extravasation zones of the fluorophore are indicated with the arrows. The arrowheads indicate cerebral blood vessels filled with FD40 in 3-month-old WT mice. The immunoreactivity to FD40 is higher in EX mice than in WT of both ages. The extravasation zones of the FD40 are larger in 3-month-old animals than in 12-month-old mice. Scale bar = 50 μm. was higher in 3-month-old EX mice in all studied brain areas compared with the control group (p < 0.0001; Fig. 5A-D). In contrast, in 12month-old mice, there was no difference between the EX and WT groups in this parameter in any of the examined structures. Comparing the two age groups, the maximal intensity of fluorescence was higher in the younger EX mice than in older animals, in all studied structures. Comparison of the maximal intensity of fluorescence between mice from the WT groups at 3-month-old and 12-month-old showed no significant changes.
The average fluorescence intensity within the extravasation zone in both 3-and 12-month-old EX mice was higher in all the studied structures compared with the control groups ( Fig. 6A-D). The results showed that this parameter was also higher in 3-month-old mice compared with the 12-month-old mice in all studied structures of EX mice.
3.5. The immunoreactivity of proteins responsible for maintaining the integrity of the functional barrier systems is higher in cortical than in subcortical vessels To assess the effect of Hch on the vascular wall we investigated the expression of Agr, MMP-9, and AQP4 in 3-month-old EX and WT mice. Agr was present in the basal lamina of the brain microvessels (Fig. 7A). MMP-9 was expressed in the endothelial cells and smooth muscles, whereas AQP4 was observed in the astrocytic perivascular endfeet (Fig. 7B). The distribution of markers was inhomogeneous in the studied brain structures. Agr and MMP-9 immunoreactivity was higher in blood vessels of the cortical structures compared with STR in both EX and WT mice (Fig. 7A). Similarly, the higher expression of AQP4 was observed mostly in cortical than subcortical vessels (Fig. 7B).

The proteins responsible for maintaining the integrity of the functional barrier systems are present in vessels of different diameter and wall thickness
The quantitative assessment of the parameters of vessels immunoreactive to Agr, MMP-9, and AQP4 was done. The results showed the expression of the above-mentioned markers in vessels of different diameters and wall thicknesses (Fig. 8). Agr predominated in larger vessels, whereas MMP-9 was observed in smaller ones. Agr was present in vessels of larger internal and external diameter and greater wall thickness in EX animals compared with the WT group. Similarly, MMP-9 was present in vessels of larger internal and external diameter and larger vascular wall thickness in EX mice compared with the WT group. AQP4 was distributed more evenly within major and minor arteries. This water channel protein was found in the perivascular zone of vessels whose diameter and wall thickness showed no difference between the EX and the WT group.

Discussion
Among the effects exerted by Hch upon cerebral vessels are increased permeability of BBB and changes in the distribution of the vascular wall markers. In this study, we investigated the effect of Hch in 3-month-old and 12-month-old mice versus the age-matched wild-type controls. The damage to the BBB was assessed based on the perivascular leakage of the fluorescently labeled dextran. The results showed that in 3-month-old Hch mice, the BBB was predominantly lesioned in the cortical vessels. This correlated with the increased expression of Agr and MMP-9. These changes were mostly present in the larger vessels of Hch mice. Therefore, the larger cortical vessels (with higher Agr and MMP-9 expression) are particularly susceptible to wall damage and plasma perivascular leakage in 3-month-old mice. This implicates the potential for brain edema development in superficial areas of the brain hemisphere supplied with the cortical arterial branches. On the other hand, it can explain the greater predisposition for metabolic impairment e.g., the flow of oxysterols through the damaged vascular wall, as well as the transmigration of inflammatory mediators or activation of microglia and astrocytes (Czuba et al., 2017). Furthermore, we did not observe a significant difference in the internal diameter of striatal vessels between the Hch and WT groups. Therefore, our observations suggest different resistance of BBB to the pathological processes in the course of Hch in the cortical vessels and deep perforating branches. This could be explained by differences in the vascular wall structure, blood flow conditions (e.g. different blood pressure in these two types of vessels) or differences in pathophysiological processes occurring in the endothelial cells of the cortical and deep brain vessels (Bogorad et al., 2019;Hainsworth et al., 2015).
The intensity of BBB damage in the course of Hch was assessed based on the diameter of the extravasation zone. In 3-month-old Hch mice, a larger extravasation zone compared with the WT group was present in all examined brain regions. In 12-month-old Hch mice, a larger Fig. 3. The internal diameter of the cerebral blood vessels injected with 40 kDa FITC-dextran in 3-month-old mice with Hch (EX) and control group (WT) in the A) prefrontal cortex (PFCx), B) motor cortex (MCx), C) hippocampus (HIP), and D) striatum (STR). In EX, BBB perivascular leakage was found in vessels of larger internal diameter in cortical structures (PFCx, MCx, and HIP), compared with the WT group. All data are presented as mean ± standard deviation (SD). Mann-Whitney U test (A-D); ****p < 0.0001 extravasation zone versus the control was present only in the HIP and STR. Age-related decrease in BBB permeability can be explained by its maturation. In 3-month-old Hch mice, the mean diameter of the extravasation zone was larger in those brain regions where the internal diameter of the vessels was larger compared with WT. The results may suggest changes in hemodynamic conditions occurring with age in the vascular system of hypercholesterolemic individuals (Hartley et al., 2000).
Apart from the assessment of the mean diameter of the extravasation zone, the study of fluorescence intensity was done. The average fluorescence intensity was higher in EX animals in both age groups, and in all examined brain regions, compared with the control. The maximal fluorescence intensity was higher within the extravasation zone of 3month-old mice versus 12-month-old mice, in all the studied brain regions. This indicates greater damage to the BBB during the maturation period (Sweeney et al., 2019). While present in both studied cortical and subcortical structures, the BBB lesion indicates the harmful effect of Hch exerted upon both morphological types of vascular branches, namely cortical and perforating arteries.
Interesting observations also result from the comparison of the morphological parameters of vessels and extravasation zone. The results suggest different predispositions to damage of the cortical and perforating arteries present in distinct brain regions. In addition, the value of the mean diameter of the extravasation zone, as well as the average and maximal fluorescence intensities of the extravasation area were higher in the adult compared with the younger animals. The high concentration of immunofluorescent dextran accumulated in the extravasation zone in the 3-month-old Hch mice may be a consequence of incomplete Fig. 4. The mean diameter of the extravasation zone after injection of 40 kDa FITC-dextran in 3-month-old and 12-month-old mice with Hch (EX) and control group (WT) assessed in the A) prefrontal cortex (PFCx), B) motor cortex (MCx), C) hippocampus (HIP), and D) striatum (STR). The mean diameter of the extravasation zone was larger in all structures examined in younger EX mice. This parameter was also larger in 3-month-old EX mice than in 12-month-old EX animals. All data are presented as mean ± standard deviation (SD). Mann-Whitney U test (A-D); **p < 0.01, ****p < 0.0001 development of the BBB and, thus, greater susceptibility to damage. Our results indicate a long-lasting BBB maturation period and differentiated BBB susceptibility to pathological processes. This is due to the development of the vascular endothelial cells, the formation of tight junctions, the maturation of specific components of the basal lamina, the vascular wall, and the perivascular space (Haddad-Tóvolli et al., 2017). It is also possible that Hch inhibits the BBB formation, extending the time to its complete maturation.
In this study, we compared the distribution of Agr, MMP-9, and AQP4 in the vascular wall of Hch and WT mice. Our results showed several differences in their localization. In both Hch and WT mice, Agr was present mainly in the basal lamina of cerebral vessels of the cortical structures (e.g. MCx and HIP). In Hch mice, this protein was present in the vessels of larger internal and external diameters and thicker arterial walls, compared with WT groups. Some previously published articles reported changes in the concentration of Agr in several brain regions (Li et al., 1997;O'Connor et al., 1994). The up-regulation of Arg can be explained by an increased blood flow in areas of the intense synaptic plasticity and the increased neuronal activity (e.g. HIP and cerebral cortex). These processes require a higher supply of oxygen and energetic metabolites, which results in the dilatation of arterioles and capillaries (Bogorad et al., 2019). This may explain our observations of uneven Agr distribution in the vessels. Considering the function of Agr, which maintains the BBB tightness and makes possible interaction between . The maximal intensity of fluorescence is higher in younger EX mice compared with the WT group and older mice in cortical and subcortical structures. All data are presented as mean ± standard deviation (SD). Mann-Whitney U test (A-D); ****p < 0.0001 astrocytes and endothelial cells (Barber and Lieth, 1997), the increase in Agr immunoreactivity in the vessels of Hch mice may be related to its protective function against BBB damage.
Although MMP-9 was abundant in the blood vessels of cortical structures, in both EX and WT animals, these vessels tend to be smaller in diameter than those immunopositive for Agr. Our results showed that in Hch mice MMP-9 was present in the vessels characterized by larger internal and external diameters and thicker arterial walls. A high MMP-9 expression was previously reported in brain areas associated with intense synaptic plasticity and the development of dendritic spines, involved in the learning and memory processes (Vafadari et al., 2016). The high expression of MMP-9 in the HIP can be due to its role in the Fig. 6. Average fluorescence intensity after injection of 40 kDa FITC-dextran in 3-month-old and 12-month-old mice with Hch (EX) and control group (WT) in A) prefrontal cortex (PFCx), B) motor cortex (MCx), C) hippocampus (HIP), and D) striatum (STR). In all examined structures of both 3-and 12-month-old EX mice, the average fluorescence intensity is higher than in the WT groups. It is also higher in 3-month-old mice than in 12-month-old animals. All data are presented as mean ± standard deviation (SD). Mann-Whitney U test (A-C, 3 M EX vs. 3 M EX group in STR, and 12 M EX vs. 12 M EX group in STR) and Student's t-test (3 M EX vs. 12 M EX group in STR); ***p < 0.010, ****p < 0.0001 Fig. 7. Microphotographs of the representative immunofluorescent staining of the vascular markers: agrin (Agr), matrix metallopeptidase 9 (MMP 9), and aquaporin-4 (AQP4) in the wall of arteries in the cerebral cortex in 3-month-old hypercholesterolemic (EX) and control (WT) mice. The investigated markers are present in various components of the vascular wall. Agr is found in the basal lamina of the brain microvessels, MMP-9 is present in the endothelial cells, whereas AQP4 is localized in the astrocytic perivascular endfeet within the perivascular space. Images were acquired on a Zeiss laser scanning confocal microscope system with ×20 objective lenses with a zoom of 2.8 times. Scale bar = 50 μm. E. Czuba-Pakuła et al. constant development of the axo-dendritic system and synaptogenesis (Ferrer-Ferrer and Dityatev, 2018), as well as the development of the hippocampal connections (Aujla and Huntley, 2014). The high expression of Agr and MMP-9 in the cortical arteries of larger diameter, observed in Hch mice, may suggest a specific role of these proteins in the development of metabolic disorders. This may be related to their greater susceptibility to BBB damage. The results of previous studies have suggested the involvement of Agr and MMP-9 in several pathological processes. For example, in AD they are involved in β-amyloid accumulation (Hayakawa et al., 2008;Ikeshima-Kataoka, 2016;Iliff et al., 2012) and changes in BBB permeability (Baumann et al., 2009;Nakada et al., 2017;Rascher et al., 2002). The increase in MMP-9 concentration has also been observed in such CNS pathologies as glutamate excitotoxicity (Michaluk and Kaczmarek, 2007), stroke (Rosell et al., 2006;Steliga et al., 2020), and inflammatory diseases (Asahi et al., 2001).
Our results showed homogeneous distribution of AQP4 in the perivascular space of all investigated brain regions, with only a slightly higher content in the cortical arteries. Previously published results showed inhomogeneous expression of AQP4 in several brain areas (Badaut et al., 2002). The highest level of AQP4 mRNA has been detected in the cerebellum and dentate gyrus (Jung et al., 1994). The low level was found in the neocortex and HIP (Vizuete et al., 1999). AQP4 is present in astrocytes and vessels in different brain areas, which confirms its important role in the regulation of water transport in brain tissue (Hoddevik et al., 2017). High levels of this water channel protein were present in hypothalamic regions involved in the preservation of osmodetection and water balance. Some authors report altered AQP4 expression associated with the development of cytotoxic edema (Wolburg et al., 2012) and glial scar formation in response to various pathological stimuli (Mader and Brimberg, 2019a;Manley et al., 2000). In our study, we did not observe major differences in AQP4 expression in the vessels between Hch and the WT groups. The lack of differences in AQP4 expression in our model can be explained by the relatively short duration of the pathological process, because our studies were carried out in 3-month-old animals. Alternatively, it could be a lack of AQP4 involvement in the pathomechanism of structural changes in the vascular wall in the course of Hch. Therefore, the effect of Hch on AQP4 expression in different brain regions awaits further investigation.
In the study, we used the low-density lipoprotein receptor and apolipoprotein double knockout (LDLR − /− /Apo E − /− ) mice, which apart from the characteristics of the LDLR − /− knockout mouse strain also exhibit the features of the Apo E − /− knockout. ApoE is an important factor involved not only in lipid metabolism but also in neurodegenerative diseases, although the role of this protein in these processes has not been extensively studied (Ferrari-Souza et al., 2023;Lou et al., 2023;Raulin et al., 2022). Under physiological conditions, ApoE is an important lipid transporter in the brain (Lou et al., 2023;Raulin et al., 2022). It participates in lipid and cholesterol metabolism, inflammatory response, maintaining synaptic transmission, and BBB integrity (Ayyubova, 2023). The apoe gene alleles encode three protein isoforms - Fig. 8. Assessment of the cerebral arteries immunostained with Agr, MMP-9, and AQP4 in 3-month-old EX and WT mice. The external A) and internal B) diameters of the vessels as well as the mean thickness of the vascular wall C) were studied. Whereas Agr and MMP-9 are present in vessels of larger internal and external diameters and a thicker vascular wall in EX mice, compared with the vessels of WT group animals, AQP4 is distributed in vessels not revealing differences in these parameters between EX and WT mice. All data are presented as mean ± standard deviation (SD). Mann-Whitney U test (A-C); ***p < 0.01, ****p < 0.0001 ApoE2, ApoE3, and ApoE4 (Raulin et al., 2022). The epsilon isoform of the apolipoprotein E (ApoEε4) allele is the strongest genetic predictor of late-onset sporadic AD (Ayyubova, 2023;de Frutos Lucas et al., 2023;Fernández-Calle et al., 2022;Kivipelto et al., 2002a). It increases the Ca 2+ ions and reactive oxygen species level in mitochondria. It also regulates the expression of genes related to aging, apoptosis, and Aβ accumulation (Pires and Rego, 2023). The ApoEε4 is associated with delayed lipoprotein catabolism, BBB damage, and an increase in the inflammatory response (D'Alonzo et al., 2023). In addition, ApoEε4 has been claimed to activate microglia in areas of the brain associated with early tau deposition in AD (Ferrari-Souza et al., 2023). To sum up, the function of various isoforms of ApoE is associated with a multidirectional impact on the development of neurodegenerative diseases (Ferrari-Souza et al., 2023), and these relationships, being only partially known, require further research.
Previous studies have been conducted in LDLR − /− knockout mice, which are considered a model of FH . Threemonth-old LDLR − /− mice develop BBB damage, especially induced by exposure to a high-cholesterol diet compared with the C57BL/6 WT mice. In our study, we have used 3-and 12-month-old LDLR − /− /Apo E − / − double knockout mice. These mice provide a useful experimental model for the study of Hch and atherosclerosis (Getz and Reardon, 2016). BBB disruption and increased permeability of vascular walls were found in both experimental models. In the FH model, these changes were observed in 3-month-old mice, whereas in our model they were present in both 3-month-old and 12-month-old mice with Hch. In both models, the changes were observed in several brain areas. Additionally, cognitive impairment was reported in the LDLR − /− mice de Oliveira et al., 2011;Mulder et al., 2004).
Reassuming, in both models of Hch the BBB disruption is associated with changes in the expression of vascular wall proteins responsible for maintaining the endothelial tight junctions. The consequence of these changes is development of the perivascular leakage, which in turn leads to the development of neuroinflammation, initiation of the cell death, and development of neurodegenerative processes. Further research is needed to confirm the presence of cognitive impairment in LDLR − / − /Apo E − /− double knockout mice. Although this issue was not addressed in our study, given the size of the extravasation zones in HIP, MCx, PFCx, and STR, as a result of BBB damage, it can be assumed that cognitive impairment may also be present in our model. However, the reduction in the size of extravasation zones that we observed in 12month-old mice allows us to hypothesize that cognitive impairment may be less pronounced in older animals. The important questions to be answered in the relatively new model used in our study relate to the characteristics of neurodegenerative changes, the occurrence of the apoptotic or necrotic cell death, and the formation of amyloid-β deposits in the course of Hch.

Conclusions
In this study, we report a greater extent of the Hch-induced BBB permeability disruption in 3-month-old mice compared with 12-monthold mice. It was present mainly in the cortical regions. Although BBB lesion and perivascular leakage occurred in both age groups, they were less developed in the older mice. Our results show different levels of expression of three studied markers of the vascular wall in the course of Hch. While in hypercholesterolemic mice mice Agr and MMP-9 were more abundant in the cortical vessels with the larger internal diameter and larger wall thickness, the expression of AQP4 in the perivascular space of arteries did not differ between Hch and control mice mice. Consequently, we conclude that the duration of Hch, animal age, and localization of the brain damage determine the extent of BBB dysfunction. Differences in the concentration and distribution of proteins responsible for maintaining the functional barrier integrity affect its permeability. The maturation of the vascular system most likely limits the progression of pathological processes and contributes to the reduction of BBB permeability. CRediT authorship contribution statement ECP: conceptualization, methodology (immunohistochemistry), investigation, writing original draft; SG: statistics and data analysis; SW: methodology (microscopy imaging), manuscript writing, review, and editing; GL: conceptualization, manuscript writing, review, and editing; MZK: methodology (animal procedures), manuscript editing; PK: conceptualization and supervision, manuscript writing, review, and editing. All authors have read and agreed to the published version of the manuscript.

Funding
This work was supported by the Ministry of Education and Science in Poland, grants no.: 01-0356/08/211,  BS.20.8.12. The funds obtained were used to purchase mice and chemical reagents necessary to carry out the experiments.

Consent to participate
Not applicable.

Consent for publication
Not applicable.

Declaration of competing interest
No conflict of interest is claimed by all Authors of this publication.

Data availability
Data will be made available on request.

Acknowledgments
The cooperation with Professor Ryszard T. Smoleński M.D., Ph.D. head of the Department of Biochemistry, Medical University of Gdańsk, providing animals for research, is kindly acknowledged. The technical assistance of Eugenia Ś wiątek in the preparation of the brain sections and Sylwia Scisłowska M.A. in the preparation of figures is greatly appreciated.
This study was conducted in cooperation with the Interdisciplinary Center for Civilization Diseases Research of the Pomeranian University Słupsk, Poland.