Methyl-Beta-Cyclodextrin Restores K IR Channel Function in Brain Endothelium of Female Alzheimer’s Disease Mice

Animal care and use

All animal care use and experimental protocols for this study were approved by the Institutional Animal Care and Use Committee of Loma Linda University and performed in accord with the National Research Council’s “Guide for the Care and Use of Laboratory Animals” (8th Edition, 2011). Experiments were performed using young Pre-AD (1–3 months; females, 31 animals) and old AD (16–19 months; females, 39 animals) 3xTg-AD mice [20] generated from homozygous breeding pairs obtained from the Jackson Laboratories [Wilmington, MA, USA; (B6;129-Tg (APPSwe,tauP301L) 1Lfa Psen1^tm1Mpm/Mmjax); Mutant Mouse Resource and Research Center (MMRRC) stock #034830]. All animals were housed on a 12:12 h light-dark cycle at 22–24°C with fresh water and food available ad libitum. For individual experiments, animals were selected randomly with respect to group and biological sex.

Solutions and reagents

Preparation and composition of all solutions have been previously described for measurements of V_m in intact endothelium freshly isolated from mouse posterior cerebral arteries [5, 22]. Briefly, physiological salt solution (PSS; pH 7.4) was prepared for continuous superfusion of cerebral endothelial tubes [(in mmol/L): 140 NaCl, 5 KCl, 2 CaCl₂, 1 MgCl₂, 10 HEPES, 10 Glucose]. For dissection of brain and arteries, PSS lacking CaCl₂ contained 0.1%bovine serum albumin. For isolation of intact endothelium, an enzyme cocktail [0.31 mg/mL papain (P4762, Sigma), 0.5 mg/mL dithioerythritol (D8255, Sigma), 0.75 mg/mL collagenase (C8051, Sigma), and 0.13 mg/mL elastase (E7885, Sigma)] was prepared in dissociation PSS containing reduced Ca²⁺ (0.1 mmol/L CaCl₂) and 0.1%bovine serum albumin. The 15 mmol/L KCl solution was prepared with an equimolar decrease of NaCl accordingly in order to maintain osmolarity. All reagents were purchased from Sigma-Aldrich (St. Louis, MO, USA) or ThermoFisher Scientific (Pittsburgh, PA, USA) unless otherwise indicated.

Dissection of cerebral arteries and isolation of endothelial tubes

Sequential dissection of mouse brain for preparation of posterior cerebral arteries and isolation of arterial endothelium have been previously described and illustrated in our previous studies [5, 22]. Briefly, following isoflurane inhalation for anesthesia and decapitation, the brain was acutely removed and placed in chilled (4°C) dissection PSS. Posterior cerebral arteries were carefully dissected from the brain and cut into segments. Arterial segments were partially digested in dissociation PSS containing the enzyme cocktail at 34°C for 10 to 12 min. Following digestion, arterial segments were triturated using borosilicate glass capillary tubes [1B100-4, World Precision Instruments (WPI), Sarasota, FL, USA] with an internal tip diameter of 80–120μm to remove adventitia, smooth muscle cells, and internal elastic lamina to isolate intact endothelial tubes. Endothelial tubes were continuously superfused with PSS at 37°C.

Recording of intracellular V _m

Recording of V_m in mouse cerebral endothelium has been previously described and illustrated [5, 22]. Briefly, V_m of isolated endothelial tubes was recorded with Axoclamp electrometers (2B and/or 900A; Molecular Devices, Sunnyvale, CA) using microelectrodes pulled (P-97, Sutter Instruments) from glass capillary tubes (GC100F-10, Warner Instruments, Holliston, MA, USA). Under continuous superfusion with PSS, an endothelial cell was penetrated with a microelectrode backfilled with 2 mol/L KCl (tip resistance: ∼150 MΩ) while viewing the endothelial tube at 400×magnification. Amplifier outputs were connected to a data acquisition system (Digidata 1550A; Molecular Devices), whereby all data were acquired at a 10 Hz frequency on a Hewlett-Packard personal computer using Axoscope software (Molecular Devices). Each experiment included a continuous intracellular V_m recording for ∼2 h.

Cholesterol assay of cerebral blood vessels

Cholesterol content was determined in total blood vessel content in the brain using Cholesterol Quantification Assay Kit (Sigma) in accord with manufacturer’s instructions. Brain microvessel fragments were isolated from the mouse brains using a previously illustrated protocol [23]. Cholesterol content (μmol/L) was normalized to total phospholipid (μmol/L) determined from the same blood vessel isolates using Phospholipid Assay Kit (Sigma) according to manufacturer’s instructions.

Cholesterol imaging of endothelial cell tubes

Endothelial cell tubes were secured on a glass coverslip with glass micropipettes using two micromanipulators and superfused with PSS while gradually raising the bath temperature to 37°C. Endothelial cell membrane cholesterol was detected by incubating cells with Filipin-III (50μg/ml, Sigma) at 37°C for ∼1.5 h under light-protected conditions. The cells were then washed with fresh PSS and images were acquired using a fluorescent microscope (Eclipse Ti-S; Nikon), a 60X objective [Plan Apo λ (numerical aperture: 0.95); Nikon)], a solid-state light source containing a violet channel (MIRA Light Engine; Lumencor, Inc., Beaverton, OR, USA), a 16-megapixel monochrome camera (DS-Qi2, Nikon), and Nikon imaging software (NIS Elements-F 4.60.00). For unpaired image acquisitions prior to staining with Filipin, endothelial cell tubes were treated with methyl-β-cyclodextrin (MβCD) for 20 min followed by 1 h washout period at 37°C. For ideal visibility, all images illustrating Filipin staining were pseudo-colored green.

Pharmacology

An effective half-maximal concentration of NS309 (EC₅₀, ∼1μmol/L; 3895, Tocris, Bristol, England) was used to directly evaluate function of SK_Ca/IK_Ca channels to produce hyperpolarization [5, 22]. As consistent with our previous studies [22], DMSO solvent (≤0.1%) for working stocks of NS309 was found to have no effect on endothelial V_m on its own for≤5 min applications used for respective drug treatments in cerebrovascular endothelial tubes. Elevated level of extracellular KCl ([K⁺]_o: 15 mmol/L) was used to stimulate endothelial K_IR2 channels for hyperpolarization [22, 24]. MβCD (1μmol/L; 332615, Sigma) was prepared in PSS to reduce cholesterol in the endothelial cell membrane [25, 26]. Respective K⁺ channel activating agents were applied before and after washout of a 20 min application of MβCD treatment. Resting V_m was typically allowed≥2 min to stabilize before application of a pharmacological agent, whereby each application was allowed sufficient time (≥3 min) to record peak V_m responses. In between individual applications, the endothelial tube was washed with PSS to baseline conditions.

Data and statistical analysis

For quantification of cholesterol levels normalized to total phospholipid content, n = 1 entails one experiment using a total blood vessel sample pooled from three mouse brains of each group (pre-AD versus AD). For V_m recordings, data analyses included resting V_m (mV) and change in V_m (ΔV_m) = peak response V_m –preceding baseline V_m. Summary data represent values averaged over 10 s during stable recordings. Filipin-III image acquisitions and V_m recordings represent the number (n) of independent experiments, each representing one endothelial tube from one mouse.

For simultaneous comparisons among both age groups±MβCD treatment, statistical analysis included a two-way Analysis of Variance (ANOVA) with Tukey’s post hoc correction for multiple comparisons. One-way ANOVA with Tukey’s post hoc examined statistical significance for respective AD groups for pre-MβCD conditions versus post-MβCD conditions at both 30 min and 60 min time points. The two-tailed unpaired t-test was used for select comparisons between two groups as pre-AD versus AD data sets (e.g., cholesterol measurements, pre- or post-MβCD treatment). Two-tailed paired student’s t-tests were used to analyze select comparisons for pre-MβCD versus post-MβCD within the Pre-AD or AD group. All graph preparation and statistical analyses were performed using GraphPad Prism (La Jolla, CA, USA). Differences between groups were accepted as statistically significant with p < 0.05. All summary data are presented as the mean±SEM. Note that the absence of statistically significant differences does not necessarily rule out physiological or pathological importance of respective study group comparisons that indicate p > 0.05.

Total cholesterol content in cerebral blood vessels during conditions of pre-AD and AD

Cholesterol dyshomeostasis is known to underlie conditions of atherosclerosis per aging [29] while the presence of the apolipoprotein E (APOE) ɛ4 allele in particular is a major risk factor for development of AD [14, 30] and associated cerebrovascular pathology [31]. Thus, we first sought to determine cholesterol levels in isolated cerebral vessels of 3xTg-AD animals during pre-AD and AD conditions. Although the mean value for the total cholesterol:phospholipid ratio is ∼40%higher for AD (0.345±0.120) versus Pre-AD (0.242±0.068), this difference is not statistically significant (p = 0.48; unpaired t-test) (Fig. 1; n = 5 independent experiments for each group, 3 animals pooled per experiment; 15 animals total per group) and should be interpreted with caution. Thus, these data suggest that cerebral blood vessels may or may not be enriched with excessive levels of cholesterol among individual animals during conditions of AD.

OPEN IN VIEWER

Fig. 1

Cerebrovascular cholesterol content among pre-AD and AD groups. Data represent molar ratio of cholesterol content per total phospholipid content for respective groups. n = number of experiment for each group (Pre-AD: 5 & Old AD: 5), whereby each group consisted of pooled purified cerebral blood vessels isolated from three female mice.

Membrane cholesterol in endothelial cell tubes during conditions of pre-AD and AD

Presence of cholesterol inclusion bodies or “crystals” appear in endothelial cells during conditions of disease such as atherogenesis [26] and can be removed using MβCD [32]. With use of Filipin to stain membrane cholesterol [26, 32], we observed fluorescent crystals in endothelial tubes during AD pathology (Fig. 2F; 15.3±0.7 crystals/mm²) that were significantly reduced in number (Fig. 2H; 3.6±0.9 crystals/mm²) following mild MβCD treatment (1μmol/L; 20 min incubation, 1 h washout) or in young, pre-AD animals regardless of MβCD treatment (Fig. 2B, D; 0.2±0.1 crystals/mm²). These data demonstrate that membrane cholesterol crystals are a feature of cerebrovascular endothelium during AD pathology which may impact function of K⁺ channels crucial to endothelial function underlying perfusion of the brain.

OPEN IN VIEWER

Fig. 2

Presence of cholesterol crystals in AD animals and reduction following MβCD treatment. Representative (A, C, E, G) brightfield and (B, D, F, H) Filipin-III images (green pseudo-color) of cerebrovascular endothelial cell tubes isolated from female animals. A, B) Pre-AD images without MβCD treatment. C, D) As shown in (A) & (B) respectively after MβCD treatment (1 mmol/L, 20 min incubation, & 1 h wash). E-H) As shown in (A), (B), (C), and (D) respectively for AD animals. White arrows indicate cholesterol “crystals” in endothelial tubes during AD conditions in the absence of MβCD treatment. The scale bar throughout panels represents 20μm. I) Summary data for number of crystals per unit area (mm²). n = number of animals and respective cerebral endothelial tubes (Pre-AD & Old AD: 4; Pre-AD & Old AD after MβCD: 3). ^*p < 0.05 (two-way ANOVA & paired t-test), Pre-AD or Old AD female Post-MβCD versus respective Pre-MβCD group; ^#p < 0.05 (two-way ANOVA & paired t-test), Old AD female Pre-MβCD versus Old AD female Post-MβCD.

Effect of MβCD on cerebrovascular endothelial SK _Ca /IK _Ca channel function

Endothelial SK_Ca/IK_Ca channels are integral to a major blood flow regulation pathway as EDH in cerebral arteries [22, 33] and arterioles [34]. We have previously found that SK_Ca/IK_Ca channel function is preserved in females while modestly enhanced (∼20%) in males during AD pathology [7]. Accordingly, we found that peak hyperpolarization to NS309 (1μmol/L) as SK_Ca/IK_Ca channel activation was not altered by advanced AD pathology or MβCD treatment (1μmol/L) (Fig. 3; p > 0.05; ANOVA or t-tests). SK_Ca/IK_Ca channel-mediated hyperpolarization was as follows across female groups: Pre-AD (Pre-MβCD, 34.2±1.2 mV & Post-MβCD, 33.0±0.7 mV; n = 11) and Old AD (Pre-MβCD, 33.1±1.6 mV & Post-MβCD, 31.1±1.0 mV; n = 12) (Fig. 3B). Altogether, these data demonstrate that cerebrovascular endothelial SK_Ca/IK_Ca channel stability during AD pathology is not altered by MβCD treatment.

OPEN IN VIEWER

Fig. 3

SK_Ca/IK_Ca channel function is not impacted by conditions of AD or membrane cholesterol regulation. A) Representative recordings of SK_Ca/IK_Ca channel hyperpolarization in Old AD females in response to NS309 (1μmol/L) before and after 30 min washout of MβCD (1 mmol/L, 20 min). B) Summary data for peak ΔV_m in response to SK_Ca/IK_Ca activation in Pre-AD versus Old AD females before and after MβCD treatment (30 min washout). n = number of animals and respective cerebral endothelial tubes (Pre-AD: 11; Old AD: 12).

Impact of MβCD treatment on cerebrovascular endothelial K _IR 2 channel function

K_IR2 channel function is crucial for cerebral vasodilation and blood flow [10, 15] and is impacted by conditions of hypercholesterolemia [16, 17]. Also, we have recently found reduced K_IR2 channel function with advancing age [5] while exacerbated by conditions of AD [7] in both biological sexes. Thus, we tested whether treating with a relatively low dose of MβCD, an agent that lowers membrane cholesterol [25, 26], could restore K_IR2 channel function in endothelial tubes from AD mice. Relative to SK_Ca/IK_Ca channel activation, there is a modest V_m hyperpolarization response to activation of cerebrovascular endothelial K_IR2 channels with 15 mmol/L KCl (ΔV_m ∼–5 to –10 mV; [5, 22]) (Fig. 4A). In female animals, we have found that hyperpolarization to 15 mmol/L KCl decreased by ∼40%(p < 0.05, unpaired t-test) from young pre-AD (7.3±0.6 mV) to old AD (4.5±0.8 mV) (Fig. 4B). As an approximate wild-type reference to the AD animals, age-matched (16–19 months) C57BL/6 females respond to 15 mmol/KCl with 7.0±1.1 mV of hyperpolarization (n = 5) [5]; similar to the young Pre-AD group. Following MβCD treatment (30 min washout), both pre-AD and AD hyperpolarization responses were similar as ∼9±1 mV (pre-AD, 9.3±1.2 mV & AD, 8.5±0.8 mV). In such manner, a statistically significant increase by ∼2-fold (p < 0.05; two-way ANOVA & paired t-test) was only observed in AD animals versus paired pre-MβCD measurements (Fig. 4A, B). Furthermore, we found that hyperpolarization responses were additionally enhanced following a 60-min washout time period after MβCD treatment, whereby hyperpolarization responses for both pre-AD (13.3±1.9 mV) and AD (13.6±1.2 mV) females were statistically significant versus respective pre-MβCD conditions (p < 0.05; One-way ANOVA). As a reference, an approximate time period of incubation with PSS (washout solution and vehicle for MβCD) did not significantly alter hyperpolarization to 15 mM KCl (n = 4 AD animals; 0 h, 4.3±0.6 mV & 1.5 h, 5.3±0.9 mV). Altogether, these data suggest that a relatively low concentration of MβCD, an agent known to reduce membrane cholesterol, can restore (with potential to further enhance) cerebrovascular endothelial K_IR2 channel function during conditions of advanced AD in females.

OPEN IN VIEWER

Fig. 4

Reduction of membrane cholesterol restores endothelial K_IR2 channel function during AD pathology. A) Representative recordings of K_IR2 channel-mediated hyperpolarization in Old AD females in response to 15 mM KCl before and after 30 min washout of MβCD (1 mmol/L, 20 min). B) Summary data for peak ΔV_m in response to K_IR2 channel activation in Pre-AD versus Old AD females before and after MβCD treatment (30 min washout). Note that both Pre-AD and Old AD females show similar hyperpolarization to 15 mM KCl as ∼9 mV following MβCD treatment. n = number of animals and respective cerebral endothelial tubes (Pre-AD: 9; Old AD: 12). ^*p < 0.05 (unpaired t-test) for Old AD Female Pre-MβCD versus Pre-AD Female Pre-MβCD; ^#p < 0.05 (two-way ANOVA & paired t-test), Old AD female Post-MβCD versus Old AD female Pre-MβCD.

Endothelial membrane cholesterol and use of cyclodextrin for treatment of AD

Cholesterol is essentially present in the plasma membrane of all mammalian cells as one of the major lipid components for cellular growth and function, whereby its excess beyond physiological levels is generally associated with the development of disease including AD [14, 44–46]. Although not statistically significant, we have found that the mean for cerebrovascular cholesterol (normalized to total phospholipid content) was ∼40%higher in AD versus pre-AD animals (Fig. 1). Thus, our data for cellular cholesterol content suggest that cerebral blood vessels may or may not be enriched with excessive levels of cholesterol among individual animals during conditions of AD. However, with use of fluorescent Filipin staining, we observed apparent cholesterol crystals in endothelial membranes of AD animals that were substantially reduced in number by ∼75%following MβCD treatment (Fig. 2). In females, elevated cholesterol levels in human frontal cortex [47] and rodent endothelial plasma membranes [48] have been shown during conditions of aging as a key risk factor for cognitive decline towards AD [1].

Cyclodextrins such as MβCD are cyclic oligosaccharide compounds and have been widely used to bind to cholesterol in a concentration-, time-, temperature-, and cell type-dependent manner [25]. With a reference point of 5 mmol/L MβCD (30 to 60 min incubation) for patch-clamp measurements of rodent vascular cells (brain [15], mesentery [16]), we determined that 1 mmol/L MβCD (20 min) was sufficient for restoration of reduced K_IR channel-dependent hyperpolarization for AD animals following 30 min of MβCD washout. In addition to the precision in time for various drug-response measurements, it should be noted that a major strength of our electrophysiological abilities encompasses a scale of hours (versus seconds or minutes) [21] and thus, we encompassed paired examinations of Pre-MβCD and Post-MβCD conditions throughout individual endothelium preparations.

SK _Ca /IK _Ca channel function is unaffected, whereas K _IR channel function is restored by mild MβCD treatment of cerebrovascular endothelial membrane cholesterol during AD

Endothelial SK_Ca/IK_Ca channels are vital for cerebral blood flow regulation [33, 34], whereby their function is not impaired by conditions of AD alone [7]. Accordingly, the current study demonstrates that MβCD has no significant impact on SK_Ca/IK_Ca channel activity to produce membrane hyperpolarization (Fig. 3), suggesting that membrane cholesterol does not play a major role for regulation of cerebrovascular SK_Ca/IK_Ca during health and disease. Overall, this lack of effect of MβCD treatment on SK_Ca/IK_Ca channel function is consistent with previous studies (reviewed in [49]). As highly expressed in smooth muscle cells [50], large conductance K_Ca (BK_Ca) channels are suppressed [51] or activated in response to membrane-enriched cholesterol depending on expression of KCNMB1 and trafficking of BK_Ca β₁ subunits [52].

K_IR channels are integral to neurovascular coupling and the overall coordination of cerebral pre-capillary vasodilation and blood flow [10, 53] with a unique capacity for also boosting SK_Ca/IK_Ca channel hyperpolarization [54, 55]. Elevated membrane cholesterol suppresses endothelial K_IR2 channel currents [16, 45] by interacting with a CD loop region of the C-terminus as a cytosolic component of the channel [56]. Remarkably, we found that K_IR2 channel-dependent hyperpolarization increased progressively with time of washout following MβCD treatment to ∼9 mV and ∼14 mV for both female Pre-AD and AD animals at 30 min and 60 min respectively (Fig. 4 and Results text). The enhancement of hyperpolarization responses in pre-AD animals was unexpected but previous evidence suggests that, even in healthy humans, normal cholesterol levels can suppress endothelial function [57]. Also, we should note that mechanisms of local control of K_IR channels can occur via glycocalyx regulation as well [43]. Relative contributions of membrane cholesterol-dependent versus -independent modulation of endothelial ion channels remain to be resolved for conditions of AD.

Experimental considerations and potential treatment strategies

Although K_IR2 channels are also present on cerebrovascular smooth muscle cells [10, 15], we studied the key component of coordinated electrical signaling of cerebral blood flow control as the endothelium [9, 22]. In addition to retention of in vivo properties as V_m regulation and intercellular electrical coupling [22], the mechanical stability of endothelial cells (versus smooth cells) for continuous, long-term electrophysiological recordings is an experimental advantage as well. It should be noted that another recent study found that cholesterol primarily regulates cerebral arterial smooth muscle versus endothelial K_IR2 channels in young (10 to 12 weeks) female rats [15]. Thus, with knowing that cholesterol can also impact smooth muscle K_IR2 channels in intact cerebral arteries [15], we believe that the current study demonstrates the general concept that mild treatment with MβCD can alleviate suppressed K_IR2 channel function during conditions of AD [7, 8]. Therapy via PIP₂ restoration may also restore capillary endothelial K_IR2 channel function for improved neurovascular coupling and cerebral blood flow during AD [8]. However, PIP₂ is also a key dynamic component of constitutive, moment-to-moment physiology (e.g., intracellular Ca²⁺ signaling), whereby its hydrolysis following activation of G_q-protein coupled receptors [58], can notably facilitate TRPV4 activation while suppressing K_IR2 channels [59]. Although such investigations of PIP₂ are mechanistically informative, we suggest that a mild reduction of membrane cholesterol may serve as a more effective strategy for improving endothelial K_IR2 channel function with significantly less impact on constitutive events governing basal cerebral blood flow.