Impaired intracellular Ca²⁺ signaling contributes to age-related cerebral small vessel disease in Col4a1 mutant mice

Abstract

Humans and mice with mutations in COL4A1 and COL4A2 manifest hallmarks of cerebral small vessel disease (cSVD). Mice with a missense mutation in Col4a1 at amino acid 1344 (Col4a1^+/G1344D) exhibit age-dependent intracerebral hemorrhages (ICHs) and brain lesions. Here, we report that this pathology was associated with the loss of myogenic vasoconstriction, an intrinsic vascular response essential for the autoregulation of cerebral blood flow. Electrophysiological analyses showed that the loss of myogenic constriction resulted from blunted pressure-induced smooth muscle cell (SMC) membrane depolarization. Furthermore, we found that dysregulation of membrane potential was associated with impaired Ca²⁺-dependent activation of large-conductance Ca²⁺-activated K⁺ (BK) and transient receptor potential melastatin 4 (TRPM4) cation channels linked to disruptions in sarcoplasmic reticulum (SR) Ca²⁺ signaling. Col4a1 mutations impair protein folding, which can cause SR stress. Treating Col4a1^+/G1344D mice with 4-phenylbutyrate, a compound that promotes the trafficking of misfolded proteins and alleviates SR stress, restored SR Ca²⁺ signaling, maintained BK and TRPM4 channel activity, prevented loss of myogenic tone, and reduced ICHs. We conclude that alterations in SR Ca²⁺ handling that impair ion channel activity result in dysregulation of SMC membrane potential and loss of myogenic tone and contribute to age-related cSVD in Col4a1^+/G1344D mice.

Introduction

Cerebral small vessel diseases (cSVDs) are a group of related pathologies that damage arteries, arterioles, venules, and capillaries in the brain. cSVDs are a substantial cause of stroke and, after Alzheimer’s disease, are the second leading cause of cognitive impairment and dementia ^1,2. The underlying pathogenic mechanisms remain largely unknown, and no specific treatment options exist. Sporadic and familial forms of cSVD share clinical and radiological features, including white matter hyperintensities, dilated perivascular spaces, lacunar infarcts, microbleeds, and intracerebral hemorrhages (ICHs) ^3–6. Individuals with Gould syndrome, a rare multisystem disorder caused by autosomal dominant mutations in collagen IV alpha 1 (COL4A1) and alpha 2 (COL4A2), manifest these hallmarks ^7–9. The majority of pathogenic COL4A1 and COL4A2 mutations result in substitutions of critical glycine (Gly) residues with other amino acids within triple helical domains ¹⁰. These mutations impair the formation and secretion of COL4A1 and COL4A2 heterotrimers [α1α1α2(IV)] leading to collagen [α1α1α2(IV)] deficiency in the basement membrane, and cause intracellular retention of misfolded mutant collagen that can lead to endoplasmic reticulum (ER) stress ^11–13. However, it is unclear how these defects lead to cSVD. Col4a1 and Col4a2 mutations demonstrate position-dependent effects contributing to phenotypic heterogeneity ¹³. For the current study, we investigated the pathogenesis of Col4a1-associated cSVD using mice heterozygous for a missense mutation that changes the Gly residue at amino acid position 1344 to aspartate (Asp). These animals, designated Col4a1^+/G1344D, model the most common type of mutation in humans and develop severe cSVD pathology compared to other mutant mice carrying glycine missense mutations in a murine allelic series of Col4a1 and Col4a2 mutations ^10,13.

Distinctive attributes of the cerebral vasculature serve the unique requirements of the brain. For example, cerebral arteries and arterioles exhibit graded constriction in response to changes in intravascular pressure. This vital autoregulatory process, known as the vascular myogenic response, protects downstream capillary beds from potentially harmful increases in perfusion pressure when cardiac output is transiently elevated during exercise or other stimuli ¹⁴. The myogenic response is intrinsic to vascular smooth muscle cells (SMCs) and requires the coordinated activity of several ion channels. Increases in intraluminal pressure activate transient receptor potential melastatin 4 (TRPM4) channels, allowing an influx of Na⁺ ions that depolarizes the plasma membrane to increase voltage-dependent Ca²⁺ inflow through Cav1.2 Ca²⁺ channels and engage the contractile apparatus ^15,16. Knockdown of TRPM4 in vivo decreases cerebral artery myogenic constriction and impairs cerebral blood flow autoregulation ¹⁷. Pressure-induced membrane depolarization is balanced by hyperpolarizing currents conducted by several K⁺-permeable ion channels, including large-conductance Ca²⁺-activated K⁺ (BK) channels, voltage-dependent K⁺ (K_V) channels, and others ^18–20. The interplay of depolarizing and hyperpolarizing currents controls the resting membrane potential of SMCs and maintains cerebral arteries in a state of partial contraction. BK and TRPM4 channels require high levels of intracellular Ca²⁺ for activation and, under native conditions, are stimulated by Ca²⁺ released from the sarcoplasmic reticulum (SR) through ryanodine receptors (RyRs) and inositol trisphosphate receptors (IP₃Rs), respectively ^18,21. Impairment of the myogenic response has been implicated in many types of cerebrovascular disease, including familial cSVDs associated with cerebral autosomal dominant arteriopathy with subcortical infarcts and leukoencephalopathy (CADASIL) caused by mutations in NOTCH3 ²². However, the impact of the Col4a1^G1344D mutation on the myogenic response is not known.

Here, we report that Col4a1^+/G1344D mice exhibit age-dependent loss of myogenic tone that is associated with spontaneous ICHs and brain lesions detected by susceptibility-weighted magnetic resonance imaging (SWI). The deficit in myogenic tone resulted from diminished stretch-induced activation of TRPM4 channels and subsequent impairment of SMC membrane depolarization. We also found that subcellular and global Ca²⁺ signals generated by releasing Ca²⁺ from the SR through RyRs and IP₃Rs were disrupted in SMCs from aged Col4a1^+/G1344D mice. Treating Col4a1^+/G1344D mice with 4-phenylbutyrate (4PBA), a small molecule with chaperone properties that improves the secretion of misfolded proteins from the ER/SR ^{10,13,23–26}, prevented SR Ca²⁺ signaling defects, preserved TRPM4 channel activity, attenuated the loss of myogenic tone, and largely prevented ICHs. These data provide evidence that impaired secretion of misfolded mutant α1α1α2(IV) collagen by SMCs from Col4a1^+/G1344D mice causes disruptions in intracellular Ca²⁺ signaling, stretch-induced activation of TRPM4 channels, pressure-induced SMC membrane potential depolarization, and myogenic constriction. Our data also suggests these defects contributed to spontaneous age-related ICHs in Col4a1^+/G1344D mice. Thus, the findings of this study reveal an age-dependent facet of cSVD pathogenesis.

Results

Col4a1^+/G1344D mice exhibit age-dependent pathology.

All studies utilized Col4a1^+/G1344D mice carrying a missense mutation at amino acid 1344 near the C-terminus of the triple helical domain (Fig. 1A) and Col4a1^+/+ littermates as controls. Age is the most critical risk factor for cSVD and vascular dementia ^27,28. To investigate age-related aspects of cSVD pathology, mice were studied at 3 and at least 12 months (M) of age, representing the young adult and middle age stages of life, respectively. Male and female mice were used throughout the study, and no sex-specific differences were detected.

Figure 1. Col4a1^+/G1344D mice exhibit age-dependent pathologies.

(A) Schematic representation of the Col4a1 point mutation that leads to Col4a1^G1344D. (B) Intracerebral hemorrhages (ICHs) were assessed using Prussian blue staining. Representative images of brain sections from 3 M-old Col4a1^+/+ and Col4a1^+/G1344D mice stained with Prussian blue. (C) Summary data presented as percentage of brain area with Prussian blue staining in brain sections (500 μm intervals). N = 3 animals per group. *P ≤ 0.05, unpaired t-test. (D) Representative images of brain sections from 12 M-old Col4a1^+/+ and Col4a1^+/G1344D mice stained with Prussian blue. (E) Summary data presented as percentage of brain area with Prussian blue staining in brain sections (500 μm intervals). N = 3 animals per group. *P ≤ 0.05, unpaired t-test. (F) Brain lesions were evaluated in vivo with high-field susceptibility-weighted magnetic resonance imaging (SWI). Representative SWI images from 12 M-old Col4a1^+/+ and Col4a1^+/G1344D mice. Red arrowheads highlight hypointense pixels that indicate hemorrhagic lesions. (G and H) Quantification of the number of lesions (G) and the total volume of lesions detected by SWI (H). N = 4 animals per group. *P ≤ 0.05, unpaired t-test. (I) Region-wise volumetric distribution of lesions detected by SWI from 12 M-old Col4a1^+/G1344D mice. Scale bars = 1 mm. Inset scale bars = 250 μm.

Spontaneous ICHs is a highly penetrant and consequential manifestation of Col4a1 mutations that increases in severity with age ^10,29,30. ICH severity was measured in brain sections from Col4a1^+/+ and Col4a1^+/G1344D mice stained with Prussian blue. ICHs were not detected in 3 M-old Col4a1^+/+ mice. Low-level, punctate Prussian blue staining was present in brain sections from 3 M-old Col4a1^+/G1344D mice (Fig. 1, B and C). 12 M-old Col4a1^+/+ mice did not exhibit ICHs, but brain sections from 12 M-old Col4a1^+/G1344D mice displayed large areas of Prussian blue staining mainly localized to the subcortical regions (Fig. 1, D and E). The total area of Prussian blue staining in 12 M-old Col4a1^+/G1344D mice was nearly 20-fold greater than in 3 M-old mutants (1.9% compared to 0.1% of brain area, respectively), indicating that the severity of spontaneous ICHs increased with age.

Neuroimaging is frequently used to diagnose cSVD in human patients ³¹. SWI showed 12 M-old Col4a1^+/+ mice did not exhibit SWI-detectable brain lesions, but brains from age-matched Col4a1^+/G1344D mice had multiple lesions, mainly localized to the thalamus (Fig. 1, F to I). The number, total volume, and localization of brain lesions detected with SWI were consistent with ICHs detected with Prussian blue staining.

Diminished pressure-induced constriction of cerebral arteries from 12 M-old Col4a1^+/G1344D mice is due to impaired SMC membrane depolarization.

To better understand the underlying causes of ICHs in Col4a1^+/G1344D mice, we used ex vivo pressure myography to study the myogenic response of cerebral arteries ³². Vasoconstriction in response to stepwise increases in intraluminal pressure (from 5 to 120 mmHg) was evaluated by measuring the active steady-state and passive luminal diameter of cannulated cerebral arteries bathed in standard physiological saline solution at each pressure and calculating myogenic tone. The myogenic tone of cerebral arteries isolated from 3 M-old Col4a1^+/+ and Col4a1^+/G1344D mice did not differ (Fig. 2, A and B). However, cerebral arteries from 12 M-old Col4a1^+/G1344D mice barely constricted in response to pressure and had significantly lower levels of myogenic tone than those from age-matched Col4a1^+/+ mice (Fig. 2, C and D). The passive diameters of cerebral arteries isolated from 12 M-old Col4a1^+/+ and Col4a1^+/G1344D mice did not differ (fig. S1), suggesting that changes in myogenic tone were not due to alterations in the mechanical properties of the vessel wall.

Figure 2. Diminished pressure-induced constriction of cerebral arteries from 12 M-old Col4a1^+/G1344D mice is due to impaired SMC membrane depolarization.

(A) Representative recordings of the inner diameter of isolated cerebral arteries from 3 M-old Col4a1^+/+ and Col4a1^+/G1344D mice show the myogenic response to increased pressure (active) and the dilation of the same arteries when extracellular Ca²⁺ has been removed (passive). (B) Summary data of myogenic tone as a function of intraluminal pressure. n = 5-10 vessels from 5 or 6 animals per group. ns = not significant, two-way ANOVA. (C) Representative recordings of the inner diameter of isolated cerebral arteries from 12 M-old Col4a1^+/+ and Col4a1^+/G1344D mice. (D) Summary data of myogenic tone as a function of intraluminal pressure. n = 13 vessels from 6 or 9 animals per group. *P ≤ 0.05, two-way ANOVA. (E) Representative membrane potential (mV) recordings of smooth muscle cells (SMCs) in pressurized cerebral arteries isolated from 3 M-old Col4a1^+/+ and Col4a1^+/G1344D mice at 20 mmHg and 80 mmHg intraluminal pressure. (F) Summary data showing membrane potential of SMCs in pressurized cerebral arteries at 20 mmHg and 80 mmHg intraluminal pressure. n = 4-5 vessels from 3 or 4 animals per group. ns = not significant, unpaired t-test. (G) Representative membrane potential recordings of SMCs in pressurized cerebral arteries isolated from 12 M-old Col4a1^+/+ and Col4a1^+/G1344D mice at 20 mmHg and 80 mmHg intraluminal pressure. (H) Summary data showing membrane potential of SMCs in pressurized cerebral arteries at 20 mmHg and 80 mmHg intraluminal pressure. n = 7-9 vessels from 5 or 6 mice per group. *P ≤ 0.05, unpaired t-test.

Increases in intraluminal pressure cause SMCs in the vascular wall to depolarize ³³, triggering voltage-dependent Ca²⁺ influx and muscular contraction ^33,34. To determine if the loss of myogenic tone in 12 M-old Col4a1^+/G1344D mice was associated with impaired pressure-induced depolarization, we recorded the membrane potentials of SMCs in intact cerebral arteries using intracellular microelectrodes. At low intraluminal pressure (20 mmHg), the membrane potential of SMCs in cerebral arteries from 3 M-old Col4a1^+/+ and Col4a1^+/G1344D mice did not differ (Fig. 2, E and F). Increasing intraluminal pressure to physiological levels (80 mmHg) depolarized the SMC membrane potential in arteries from both groups to the same extent (Fig. 2, E and F). SMC membrane potential did not differ between vessels from 12-M old Col4a1^+/+ or Col4a1^+/G1344D animals when arteries were pressurized to 20 mmHg, but when intraluminal pressure was increased to 80 mmHg, the depolarization of SMCs in cerebral arteries from Col4a1^+/G1344D mice was significantly blunted compared with controls (Fig. 2, G and H). These data suggest that cerebral arteries from 12 M-old Col4a1^+/G1344D mice lose myogenic tone because pressure-induced SMC membrane depolarization is impaired.

Age-dependent declines in BK and TRPM4 channel activity in SMCs from Col4a1^+/G1344D mice.

To determine if changes in ion channel activity account for impaired pressure-induced SMC depolarization, we used patch-clamp electrophysiology to compare hyperpolarizing K_V and BK currents and depolarizing TRPM4 currents in native SMCs from control and Col4a1^+/G1344D mice. Elevated SMC K_V channel current density is responsible for diminished myogenic tone in cerebral arteries from CADASIL cSVD mice ²². We measured K_V currents in freshly isolated SMCs to determine if a similar mechanism accounts for the loss of myogenic tone in cerebral arteries from 12 M-old Col4a1^+/G1344D mice. K_V currents were recorded in the perforated patch-clamp configuration to maintain the integrity of intracellular signaling cascades. Whole-cell currents were evoked by applying voltage steps from a holding potential of −80 mV in the presence of selective BK channel blocker paxilline, then in the presence of K_V channel blocker 4-aminopyridine. K_V current amplitude was determined at each potential by subtraction. We found that K_V current amplitudes did not differ between SMCs from 12 M-old Col4a1^+/+ and Col4a1^+/G1344D mice at all applied command potentials (fig. S2, A and B).

The activity of BK channels in cerebral artery SMCs under native conditions is driven by the transient release of Ca²⁺ ions through RyRs on the SR into restricted subcellular microdomains proximal to the plasma membrane ³⁵. These signaling events, known as Ca²⁺ sparks, activate clusters of BK channels on the plasma membrane to generate large-amplitude spontaneous transient outward currents (STOCs) ^18,19. We utilized the perforated patch-clamp configuration to measure STOCs in cerebral artery SMCs from 3 M-old Col4a1^+/+ and Col4a1^+/G1344D mice over a range of command potentials (−60 to 0 mV) and saw no significant difference in STOC frequency or amplitude (Fig. 3, A to C). However, the frequency and amplitude of STOCs were significantly lower in cerebral artery SMCs from 12 M-old Col4a1^+/G1344D mice compared with age-matched controls (Fig. 3, D to F). To determine if decreases in STOCs resulted from changes in the amount of functional BK channels at the plasma membrane, we utilized the conventional whole-cell patch clamp configuration to measure K⁺ currents during voltage-clamp steps from a holding potential of −80 mV. The BK-dependent component of the current was isolated with paxilline. We found that BK current amplitudes did not differ between SMCs from 12 M-old Col4a1^+/+ and Col4a1^+/G1344D mice at all applied command potentials (fig. S3, A and B). We also assessed gene expression by ddPCR to measure transcript levels of Kcnma1 (which encodes the BK α subunit), Kcnmb1 (which encodes the BK β1 subunit), and Ryr2 (which encodes ryanodine receptor 2). No differences in mRNA expression levels were detected (fig S4). Because reduced BK channel activity depolarizes SMC membrane potential and increases contractility ³⁶, diminished BK channel activity in SMCs from 12 M-old Col4a1^+/G1344D mice cannot account for impaired pressure-induced membrane depolarization.

Figure 3. Age-dependent declines in BK and TRPM4 channel activity in SMCs from Col4a1^+/G1344D mice.

(A) Representative traces of spontaneous transient outward currents (STOCs) in freshly isolated cerebral artery smooth muscle cells (SMCs) from 3 M-old Col4a1^+/+ and Col4a1^+/G1344D mice over a range of membrane potentials (−60 to 0 mV). (B and C) Summary data showing STOC frequency (B) and amplitude (C) at each command potential. n = 8-12 cells from 5 animals per group. ns = not significant, two-way ANOVA. (D) Representative traces of STOCs in cerebral artery SMCs from 12 M-old Col4a1^+/+ and Col4a1^+/G1344D mice. (E and F) Summary data showing STOC frequency (E) and amplitude (F) at each command potential. n = 12 cells from 5 animals per group. *P ≤ 0.05, two-way ANOVA. (G) Representative I-V plots of whole-cell patch-clamp TRPM4 current recordings during voltage ramps (−100 to 100 mV) in cerebral artery SMCs from 3 M-old Col4a1^+/+ and Col4a1^+/G1344D mice. Currents were evoked by including 200 μM free Ca²⁺ in the intracellular solution, and the TRPM4 portion of the current was determined by adding the TRPM4 blocker 9-phenanthrol (30 μM). (H) Summary of whole-cell TRPM4 current density at +100 mV. n = 8-10 cells from 3 or 4 animals per group. ns = not significant, unpaired t-test. (I) Representative I-V plots of whole-cell patch-clamp TRPM4 current recordings in cerebral artery SMCs from 12 M-old Col4a1^+/+ and Col4a1^+/G1344D mice. (J) Summary of whole-cell TRPM4 current density at +100 mV. n = 10-12 cells from 4 animals per group. ns = not significant, unpaired t-test. (K) Representative traces of transient inward cation currents (TICCs) evoked by applying negative pressure (−20 mmHg) to stretch the plasma membrane of voltage-clamped (−70 mV) cerebral artery SMCs isolated from 3 M-old Col4a1^+/+ and Col4a1^+/G1344D mice. (L) Summary data of TICC activity. n = 5-6 cells from 3 animals per group. ns = not significant, unpaired t-test. (M) Representative traces of TICCs in cerebral artery SMCs isolated from 12 M-old Col4a1^+/+ and Col4a1^+/G1344D mice. (N) Summary data of TICC activity. n = 8-10 cells from 4 or 5 animals per group. *P ≤ 0.05, unpaired t-test.

TRPM4 is a Ca²⁺-activated, nonselective monovalent cation channel required for pressure-induced SMC depolarization and the development of myogenic tone in cerebral arteries ^15,16,37,38. At the membrane potentials of SMCs in the vascular wall under physiological conditions (−60 to −30 mV), TRPM4 channels conduct inward Na⁺ currents that depolarize the plasma membrane in response to increases in intraluminal pressure ^15,39. Impaired TRPM4 channel activity is responsible for loss of myogenic tone in a mouse model of Gould syndrome caused by a Col4a1^G394V mutation ³⁰. Therefore, patch-clamp electrophysiology was used to determine if TRPM4 activity was diminished in Col4a1^+/G1344D mice. We assessed channel function and availability in freshly isolated cerebral artery SMCs using the conventional whole-cell patch-clamp configuration with a high [Ca²⁺] intracellular solution to directly activate TRPM4. Currents were recorded as voltage ramps were applied in the absence or presence of theTRPM4 blocker 9-phenanthrol¹⁵, and TRPM4 currents were isolated by subtraction. Prior studies using ion substitution protocols have shown that Ca²⁺-activated, 9-phenanthrol currents recorded using these methods are carried by Na⁺ ions, providing substantial evidence that these are bona fide TRPM4 currents ^16,37,40. We found that the amplitudes of whole-cell TRPM4 currents recorded from cerebral artery SMCs did not differ between Col4a1^+/+ and Col4a1^+/G1344D mice at 3 and 12 M of age (Fig. 3, G to J). These data indicate that the collagen mutation studied here does not directly affect TRPM4 channel function or the trafficking of channel protein to the plasma membrane.

TRPM4 channels in SMCs are activated by a signal transduction cascade initiated by angiotensin II type 1 receptors (AT1Rs) ^41–44. Activation of Gα_q protein-coupled AT1Rs stimulates phospholipase C (PLC), leading to the cleavage of phosphatidylinositol 4,5-bisphosphate (PIP₂) into diacylglycerol and inositol triphosphate (IP₃). IP₃ stimulates Ca²⁺ release from the SR through IP₃Rs to transiently activate TRPM4 channels ^21,38,41. AT1Rs are directly activated by the stretch of the plasma membrane (independently of their ligand) and are a critical mechanosensor in SMCs ^41–44. To measure the activity of TRPM4 channels in response to mechanical force, we patch-clamped SMCs in the perforated patch configuration and applied negative pressure (−20 mmHg) through the patch pipet to stretch the plasma membrane. Stretch-induced transient inward cation currents (TICCs) evoked in this manner are blocked by 9-phenanthrol or TRPM4 knockdown ^15,21. Stretch-activated TICC activity in SMCs from 3 M-old Col4a1^+/+ and Col4a1^+/G1344D (Fig. 3, K and L) did not differ. In contrast, stretch-induced TICC activity in SMCs from 12 M-old Col4a1^+/G1344D mice was significantly lower than that of age-matched Col4a1^+/+ mice (Fig. 3, M and N). To determine if altered gene expression underlies the reduction in TICC activity, we utilized ddPCR to measure transcript levels of Trpm4 (which encodes TRPM4), Itpr1 (which encodes IP₃R1), and Itpr2 (which encodes IP₃R2) in cerebral arteries from 12 M-old Col4a1^+/+ and Col4a1^+/G1344D mice. No differences in mRNA copy numbers were observed (fig. S4). These data suggest that blunted pressure-induced SMC depolarization and impaired myogenic tone result from diminished stretch-induced activation of TRPM4 channels.

Defective intracellular Ca²⁺ signaling in cerebral artery SMCs from 12 M-old Col4a1^+/G1344D mice.

Ca²⁺ released from the SR drives BK and TRPM4 channel activity in cerebral artery SMCs ^18,21. Therefore, we investigated the possibility that these vital Ca²⁺ signaling pathways are disrupted in SMCs from 12 M-old Col4a1^+/G1344D mice. High-speed, high-resolution spinning disk confocal microscopy imaging of freshly isolated SMCs loaded with the Ca²⁺-sensitive fluorophore Fluo-4 AM detected spontaneous Ca²⁺ sparks in SMCs isolated from 3 M-old Col4a1^+/+ and Col4a1^+/G1344D mice. The frequency, amplitude, duration, rise time, and decay rate did not differ between groups (Fig. 4, A to C). Spontaneous Ca²⁺ sparks were also present in SMCs isolated from both groups of 12 M-old mice. No significant differences in Ca²⁺ spark amplitude, duration, rise time, and decay rate were observed, but Ca²⁺ spark frequency was significantly lower in SMCs from 12 M-old Col4a1^+/G1344D mice (Fig. 4, D to F). Thus, diminished Ca²⁺ spark frequency likely accounts for decreased STOC frequency in SMCs from 12 M-old Col4a1 mutant mice.

Figure 4. Defective intracellular Ca²⁺ signaling in cerebral artery SMCs from 12 M-old Col4a1^+/G1344D mice.

(A) Representative time-course spinning disk confocal images exhibiting Ca²⁺ spark events in Fluo-4-AM loaded cerebral artery smooth muscle cells (SMCs) from 3 M-old Col4a1^+/+ and Col4a1^+/G1344D mice. Time = milliseconds, scale bar = 5 μm. (B) Representative traces of Ca²⁺ spark events in cerebral artery SMCs isolated from 3 M-old Col4a1^+/+ and Col4a1^+/G1344D mice presented as changes in fractional fluorescence (ΔF/F₀) over time. (C) Summary data showing Ca²⁺ spark frequency, amplitude, duration, rise time, and decay. n = 4-6 cells from 3 animals per group. ns = not significant, unpaired t-test. (D) Representative time-course images exhibiting Ca²⁺ spark events in cerebral artery SMCs from 12 M-old Col4a1^+/+ and Col4a1^+/G1344D. Time = milliseconds, scale bar = 5 μm. (E) Representative traces of Ca²⁺ spark events in cerebral artery SMCs isolated from 12 M-old Col4a1^+/+ and Col4a1^+/G1344D mice presented as ΔF/F₀ over time. (F) Summary data showing Ca²⁺ spark frequency, amplitude, duration, rise time, and decay. n = 7-8 cells from 3 or 4 animals per group. *P ≤ 0.05, ns = not significant, unpaired t-test. (G) Representative traces showing whole-cell F/F₀ changes in response to caffeine (10 mM) in Fluo-4-AM loaded cerebral artery SMCs from 3 M-old Col4a1^+/+ and Col4a1^+/G1344D mice. (H) Summary data of the ΔF/F₀ in response to caffeine. n = 8-9 cells from 3 animals per group. ns = not significant, unpaired t-test. (I) Representative traces showing whole-cell F/F₀ changes in response to caffeine in SMCs from 12 M-old Col4a1^+/+ and Col4a1^+/G1344D mice. (J) Summary data of the ΔF/F₀ in response to caffeine. n = 6-7 cells from 3 animals per group. *P ≤ 0.05, unpaired t-test. (K) Representative traces showing whole-cell F/F₀ changes in response to U46619 (100 nM) in Fluo-4-AM loaded cerebral artery SMCs from 3 M-old Col4a1^+/+ and Col4a1^+/G1344D mice. (L) Summary data of the ΔF/F₀ in response to U46619. n = 8-9 cells from 3 animals per group. ns = not significant, unpaired t-test. (M) Representative traces showing whole-cell F/F₀ changes in response to U46619 in SMCs from 12 M-old Col4a1^+/+ and Col4a1^+/G1344D mice. (N) Summary data of the ΔF/F₀ in response to U46619. n = 6-7 cells from 3 animals per group. *P ≤ 0.05, unpaired t-test.

Fluo-4 AM-loaded SMCs were challenged with the RyR agonist caffeine to investigate how RyR function was affected by the Col4a1^G1344D mutation ⁴⁵. Ca²⁺ signals arising from intracellular release were isolated from Ca²⁺ influx events by acutely removing Ca²⁺ from the extracellular solution immediately before recording. Caffeine-evoked global cytosolic [Ca²⁺] increases did not differ between SMCs from 3 M-old Col4a1^+/+ and Col4a1^+/G1344D mice (Fig. 4, G and H). In contrast, caffeine-induced global Ca²⁺ signals were smaller in amplitude in SMCs from 12 M-old Col4a1^+/G1344D mice than in controls (Fig. 4, I and J). These data indicate that the release of SR Ca²⁺ from RyRs is impaired in the SMCs of Col4a1^+/G1344D mice in an age-dependent manner.

Fluo-4 AM loaded-SMCs were also treated with the Gα_q protein-coupled thromboxane A2 receptor agonist U46619 to investigate IP₃R-mediated Ca²⁺ signaling. U46619-induced Ca²⁺ signals did not differ between cerebral artery SMCs from 3 M-old Col4a1^+/+ and Col4a1^+/G1344D mice (Fig. 4, K and L). In contrast, Ca²⁺ signals induced by U46619 were smaller in amplitude in SMCs from 12 M-old Col4a1^+/G1344D mice compared to controls (Fig. 4, M and N). These data suggest that the release of SR Ca²⁺ from IP₃Rs is also impaired in the SMCs of Col4a1^+/G1344D mice in an age-dependent manner.

RyRs and IP₃Rs share a common SR Ca²⁺ pool ⁴⁶. Impairment of Ca²⁺ release through both receptors suggests that the SR [Ca²⁺] of SMCs from 12 M-old Col4a1^+/G1344D mice is lower than that of corresponding controls. Further, diminished SR Ca²⁺ levels likely account for decreased Ca²⁺ spark frequency and diminished STOC and TICC activity.

4PBA prevents age-related cerebral artery defects in Col4a1^+/G1344D mice.

Col4a1 mutations impair collagen α1α1α2(IV) secretion in vitro and in vivo, which can be alleviated by treatment with 4PBA ^10,13,26,47. The chemical chaperone 4PBA facilitates the trafficking of misfolded mutant proteins and reduces ER/SR stress ⁴⁸. To investigate a potential link between impaired collagen secretion and the cerebral vascular pathology of Col4a1^+/G1344D mice, 4PBA was added to the drinking water of mutant and control animals from birth (postnatal day 0).

We found that 4PBA treatment prevented the diminishment of Ca²⁺ signals evoked by caffeine (Fig. 5, A and B) or U46619 (Fig. 5, C and D) in cerebral artery SMCs from 12 M-old Col4a1^+/G1344D mice. 4PBA treatment also prevented age-dependent declines in Ca²⁺ spark frequency (fig. S5, A to C). We also found that 4PBA treatment averted the loss of STOC (fig. S5, D to F) and TICC (Fig. 5, E and F) activity. These data demonstrate that age-related defects in intracellular Ca²⁺ signaling that control ion channel activity and membrane potential were prevented by treating Col4a1^+/G1344D mice with 4PBA. 4PBA treatment also prevented loss of myogenic tone (Fig. 5, G and H) and significantly diminished spontaneous ICHs in 12 M-old Col4a1^+/G1344D mice (Fig. 5, I and J). These data suggest that 4PBA protects against age-dependent defects in myogenic tone and ICHs in Col4a1^+/G1344D mice.

Figure 5. 4PBA prevents age-related cerebral artery defects in Col4a1^+/G1344D mice.

(A) Representative traces showing whole-cell F/F₀ changes in response to caffeine (10 mM) in Fluo-4-AM loaded cerebral artery SMCs from 4PBA treated 12 M-old Col4a1^+/+ and Col4a1^+/G1344D and untreated Col4a1^+/G1344D mice. (B) Summary data of ΔF/F₀ in response to caffeine. n = 6-12 cells from 3-4 animals per group. *P ≤ 0.05, ns = not significant, one-way ANOVA followed by Tukey’s multiple comparisons post-hoc test. (C) Representative traces showing whole-cell F/F₀ changes in response to U46619 (100 nM) in Fluo-4-AM loaded cerebral artery SMCs from 4PBA treated 12 M-old Col4a1^+/+ and Col4a1^+/G1344D and untreated Col4a1^+/G1344D mice. (D) Summary data of ΔF/F₀ in response to U46619. n = 4-6 cells from 3-4 animals per group. *P ≤ 0.05, ns = not significant, one-way ANOVA followed by Tukey’s multiple comparisons post-hoc test. (E) Representative traces of transient inward cation currents (TICCs) evoked by applying negative pressure (−20 mmHg) to stretch the plasma membrane of voltage-clamped (−70 mV) cerebral artery SMCs isolated from 4PBA treated 12 M-old Col4a1^+/+ and Col4a1^+/G1344D and untreated Col4a1^+/G1344D mice. (F) Summary data of TICC activity. n = 7-11 cells from 3-4 animals per group. *P ≤ 0.05, ns = not significant, one-way ANOVA followed by Tukey’s multiple comparisons post-hoc test. (G) Representative recordings of the inner diameter of isolated cerebral arteries from 4PBA-treated 12 M-old Col4a1^+/+ and Col4a1^+/G1344D and untreated Col4a1^+/G1344D mice showing the myogenic response to increases in pressure (active) and the dilation of the same arteries when extracellular Ca²⁺ has been removed (passive). (H) Summary data of myogenic tone as a function of intraluminal pressure. n = 6 or 9 vessels from 3 or 5 mice. *P ≤ 0.05, ns = not significant, two-way ANOVA followed by Sidak’s multiple comparisons post-hoc test. (I) Representative images of Prussian blue stained brain sections from 4PBA treated 12 M-old Col4a1^+/+ and Col4a1^+/G1344D and untreated Col4a1^+/G1344D mice. Scale bar = 1 mm. (J) Summary data presented as percentage of brain area with Prussian blue staining in brain sections (500 μm intervals). N = 3 animals per group. *P ≤ 0.05, ns = not significant, one-way ANOVA followed by Tukey’s multiple comparisons post-hoc test.

Discussion

Despite the devastating consequences of the disease, the molecular pathogenesis of cSVD is essentially unknown. Here we utilized the Col4a1^+/G1344D mouse model of Gould syndrome to investigate how this monogenic form of cSVD damages small arteries in the brain during aging. Our findings showed that these mice exhibited pathological changes with age, including the loss of myogenic tone, high levels of spontaneous ICHs, and brain lesions. Electrophysiology experiments revealed that the loss of myogenic tone was associated with diminished pressure-induced SMC depolarization and decreased activities of BK and TRPM4 channels. In addition, we found that fundamental RyR- and IP₃R-dependent Ca²⁺ signaling pathways were disrupted in SMCs from 12 M-old Col4a1^+/G1344D mice, likely due to a reduction in SR Ca²⁺ levels. Treating Col4a1^+/G1344D mice with the chemical chaperone 4PBA negated Ca²⁺ signaling disruptions, increased BK and TRPM4 channel activity, prevented loss of myogenic tone, and reduced spontaneous ICHs. We conclude that the age-dependent cSVD pathogenesis of Col4a1^+/G1344D mice results from defects in SMC SR function that lead to reduced Ca²⁺ store load and disrupted intracellular Ca²⁺ signaling. Defects in subcellular Ca²⁺ signals diminish BK and TRPM4 channel activity, disrupting membrane potential regulation and impeding pressure-induced constriction of cerebral arteries. These findings reveal a mechanism that contributes to age-dependent cSVD (Fig. 6).

Figure 6. Proposed mechanism for Col4a1^G1344D-linked cSVD.

Impaired trafficking of mutant collagen IV in the sarcoplasmic reticulum (SR) of cerebral artery smooth muscle cells (SMCs) causes a decrease in SR Ca²⁺. Reduction of SR Ca²⁺ disrupts Ca²⁺ signaling between inositol trisphosphate receptors (IP₃Rs) and transient receptor potential melastatin 4 (TRPM4) channels. Decreased TRPM4 channel activity impairs pressure-induced depolarization and decreases voltage-dependent Ca²⁺ influx through Ca_V1.2 Ca²⁺ channels causing a decrease in SMC contractility. Impaired cerebral artery SMC contraction causes a loss of myogenic tone that increases susceptibility to intracerebral hemorrhage. Treating mice with 4-phenylbutyrate (4PBA), a small molecule with chaperone properties, promotes the secretion of mutant collagen IV, thereby preventing reduction of SR Ca²⁺ and associated downstream defects.

Myogenic constriction protects delicate capillaries in the brain from pressure overload during increases in cardiac output and perfusion associated with exercise, stress, and other stimuli ¹⁴. Intracerebral hemorrhages in 12 M-old Col4a1^+/G1344D mice were predominantly localized to the thalamus. This may be due to the proximity of the arteries that supply the thalamus to the Circle of Willis. The paramedian thalamic and tuberothalamic arteries branch directly off the Circle of Willis and the thalamogeniculate and posterior choroidal arteries branch off the posterior cerebral arteries proximal to the Circle of Willis ⁴⁹. The short distance from large arteries that carry higher pressures could account for the increased risk of intracerebral hemorrhage in the thalamus. The myogenic response is deficient in several monogenetic forms of cSVD ^22,30,50. However, the defects responsible for the loss of myogenic tone differ between disease models. For example, Dabertrand et al. linked the loss of myogenic tone in the CADASIL model with blunted pressure-induced SMC membrane potential depolarization and connected the defect in membrane potential regulation with increased surface expression of K_V channels and elevated hyperpolarizing K⁺ current density ²². Notably, depletion of the minor membrane phospholipid PIP₂, leading to the loss of TRPM4 channel activity, is the fundamental defect responsible for the loss of myogenic constriction in a mouse model of Gould syndrome caused by a Col4a1^G394V mutation ³⁰. Although both Col4a1^G394V and Col4a1^G1344D mutations cause TRPM4-related cerebrovascular dysfunction, there is no evidence of decreased PIP₂ levels in Col4a1^+/G1344D mice. The current study uncovered a distinctive flaw in the SMCs of Col4a1^+/G1344D cSVD mice – impaired Ca²⁺ signaling. Our data showed that the release of Ca²⁺ from the SR through RyRs and IP₃Rs became faulty in Col4a1^+/G1344D mice by 12 M of age. Impaired SR Ca²⁺ release disrupted the activity of BK and TRPM4 channels. These channels play opposing roles in regulating SMC membrane potential and contractility. TRPM4 channels depolarize the membrane and promote contractility, whereas BK channels provide negative feedback to limit the magnitude of this response ⁵¹. When pressure-induced membrane depolarization is impaired, the BK pathway is disengaged and has minimal influence on membrane potential ⁵².

Diminished levels of Ca²⁺ in the SR likely cause Ca²⁺ signaling and ion channel defects in SMCs from 12 M-old Col4a1^+/G1344D mice. This concept is supported by prior studies showing that a ~16% decrease in SR [Ca²⁺] diminished STOC frequency by 70% and that a >20% decrease in SR [Ca²⁺] completely abolished IP₃R-mediated Ca²⁺ release in native SMCs ^53,54. In many types of cells, store-operated Ca²⁺ entry (SOCE) protects against decreases in ER/SR [Ca²⁺] ⁵⁵. However, SOCE is virtually undetectable in native, contractile vascular SMCs, thereby enhancing their susceptibility to pathogenic SR Ca²⁺ depletion ^52,56,57. We conclude that SR Ca²⁺ store depletion and the resulting dysregulation of BK, TRPM4, and SMC membrane potential is responsible for the age-related loss of myogenic tone in Col4a1^+/G1344D cSVD mice.

Steady-state free Ca²⁺ levels in the ER/SR are maintained by the release of Ca²⁺ into the cytosol through RyRs and IP₃Rs, mobile and stationary Ca²⁺-binding proteins in the SR lumen, and transport of Ca²⁺ from the cytosol into the SR by sarco/endoplasmic reticulum ATPase (SERCA) enzymes ⁴⁶. Disruption of one or more of these elements could lead to SR Ca²⁺ defects in Col4a1^+/G1344D mice. These mice display impaired secretion and increased intracellular accumulation of collagen α1α1α2(IV), presumably due to difficulties associated with the trafficking of misfolding of mutant proteins out of the ER/SR ^10,13. Misfolded proteins can cause ER/SR stress ⁵⁸ that is associated with the depletion of Ca²⁺ stores ⁵⁹. For example, Yamato et al. ⁶⁰ demonstrated that tunicamycin-induced ER stress promotes RyR-mediated Ca²⁺ leak, potentially due to increased reactive oxygen species (ROS) production. Our data showed that treatment with a molecular chaperone prevented SR Ca²⁺ handling defects and loss of myogenic tone and substantially attenuated spontaneous ICHs. These data provide evidence that impaired secretion of mutant collagen IV contributes to SMC Ca²⁺ signaling defects and cSVD pathology of Col4a1^+/G1344D mice. The age dependency of the disease process may reflect the normal functional decline in SR protein homeostasis ⁶¹ that is exacerbated by mutant collagen in the SR. A systematic review of patients with COL4A1 mutations revealed that the first presentation of the disease had a mean onset age of 36 years old ⁹. This would suggest humans between 20-30 years old are able to compensate similarly to 3-month-old mice, but the ability to regulate misfolded proteins declines with age leading to increased SR stress.

We conclude that age-dependent disruption of Ca²⁺ signaling in SMCs due to impaired secretion of mutant collagen and diminished SR Ca²⁺ store load contributes to cSVD pathology in Col4a1^+/G1344D mice. Future investigations into links between SMC SR stress and Ca²⁺ handling defects in other genetic and sporadic forms of cSVD may provide insight into novel pathogenic mechanisms.

Materials and Methods

Chemical and reagents:

Chemicals and other reagents were obtained from Sigma-Aldrich, Inc. (St. Louis, MO, USA) unless otherwise specified.

Animals:

Young adult (~3M) and middle-aged (~12M) male and female littermate Col4a1^+/+ and Col4a1^+/G1344D mice were used in this study. The Col4a1^G1344D mutation was backcrossed to C57BL/6J mice for over 20 generations ⁶². Animals were maintained in individually ventilated cages (≤5 mice/cage) with ad libitum access to food and water in a room with controlled 12-hour light and dark cycles. All animal care procedures and experimental protocols involving animals complied with the National Institutes of Health (NIH) Guide for the Care and Use of Laboratory Animals and were approved by the Institutional Animal Care and Use Committees at the University of Nevada, Reno, and University of California, San Francisco.

4PBA treatment:

Mice were provided with 4-phenylbutyrate (4PBA; 50 mM; Scandinavian Formulas Inc., Sellersville, PA, USA) from birth in drinking water that was refreshed weekly as described previously ²⁶.

Histological analysis:

Isoflurane anesthetized mice were transcardially perfused with ice-cold phosphate-buffered saline (PBS) followed by ice cold 4% paraformaldehyde (Fisher, Waltham, MA, USA) in PBS. Brains were postfixed with 4% paraformaldehyde for 24 hours at 4°C, cryoprotected in 30% sucrose in PBS, and embedded in optimal cutting temperature compound (Fisher). Coronal cryosections (35 μm) regularly spaced (500 μm) along the rostrocaudal axis were stained with a Prussian blue and nuclear fast red stain kit (Abcam, Cambridge, UK) following the manufacturer’s protocol. Images were acquired with a BZ-X700 microscope using BZ-X Viewer 1.3.0.5 software and stitched with BZ-X Analyzer 1.3.0.3 software (Keyence, Osaka, Japan). The percentage of brain area with Prussian blue staining for each section was calculated with ImageJ software (v2.3.0/1.53f; NIH, Bethesda, MD, USA). Hemorrhage severity was expressed as the percentage of hemosiderin surface area on ~21 sections for each brain (%) = [area of Prussian blue staining/total area of brain slices)] × 100.

In vivo magnetic resonance imaging:

In vivo magnetic resonance (MR) imaging was conducted on a 14.1 Tesla vertical MR system (Agilent Technologies, Palo-Alto, CA) equipped with 100 G/cm gradients and a single tuned millipede ¹H proton coil (∅_I = 40 mm). Before imaging, mice were anesthetized using isoflurane (1-1.5% in O₂), positioned in a dedicated cradle maintaining constant anesthesia, and placed in the MR bore; respiration and temperature were continuously monitored during all acquisitions to ensure animal well-being and data reproducibility. Susceptibility weighted imaging (SWI) was performed to acquire high-resolution images of the mouse brain using the following parameters: gradient-echo scheme, field of view (FOV) = 20x 20 mm², matrix = 256 x 256, 16 slices, 0.4 mm slice thickness, 0.1 mm interslice gap, number of averages = 16, echo time (TE)/repetition time (TR) = 4.60/140ms, flip angle = 10 degrees. T2-weighted (T2W) images were also acquired in a fast-spin-echo scheme to evaluate anatomical brain structures, using the same FOV geometry as SWI and the following parameters: number of averages = 8, TE/TR = 21.38/2500ms, flip angle = 90 degrees. The fast-spin-echo based T2W imaging is less sensitive to distortion artifacts caused by B₀ inhomogeneity than SWI and is thus used for the atlas-based brain structure registration as described below.

Using in-house software, hypointense lesions corresponding to ICHs were quantified based on manual segmentation of the SWI images. To investigate the number and distribution of hypointense SWI lesions across the brain, an open-source atlas-based imaging data analysis pipeline (AIDAmri) ⁶³ was customized to register T2W brain images to the Allen Brain Reference Atlas ⁶⁴. Based on the registration result, the normalized lesion volumes for each brain region were summed up across all animals and an intra-group region-wise volumetric analysis was performed.

Isolation of cerebral arteries and SMCs:

Mice were euthanized by decapitation under isoflurane anesthesia. Cerebral arteries were isolated from the brain in Ca²⁺-free Mg²⁺ physiological saline solution (Mg²⁺-PSS) consisting of 140 mM NaCl, 5 mM KCl, 2 mM MgCl₂, 10 mM HEPES, and 10 mM glucose (pH 7.4, adjusted with NaOH), with 0.5% bovine serum albumin (BSA). Native SMCs for patch-clamp experiments were obtained by initially digesting isolated arteries in 1 mg/mL papain (Worthington Biochemical Corporation, Lakewood, NJ, USA), 1 mg/mL dithiothreitol (DTT), and 10 mg/mL BSA in Ca²⁺-free PSS at 37°C for 12 minutes, then incubating with 1 mg/mL type II collagenase (Worthington Biochemical Corporation) for 14 minutes. A single-cell suspension was prepared by washing digested arteries three times with Mg²⁺-PSS and triturating with a fire-polished glass pipette. Cells used for this study were freshly dissociated on the day of experimentation.

Pressure myography:

The current best practice guidelines for pressure myography experiments were followed ³². Arteries were mounted between two glass cannulas (outer diameter, 40–50 μm) in a pressure myograph chamber (Living Systems Instrumentation, St Albans City, VT, USA) and secured by a nylon thread. Intraluminal pressure was controlled with a servo-controlled peristaltic pump (Living Systems Instrumentation), and preparations were visualized with an inverted microscope (Accu-Scope Inc., Commack, NY, USA) coupled to a USB camera (The Imaging Source LLC, Charlotte, NC, USA). Changes in luminal diameter were assessed using IonWizard software (version 7.2.7.138; IonOptix LLC, Westwood, MA, USA). Arteries were bathed in warmed (37°C), oxygenated (21% O₂, 6% CO₂, 73% N₂) PSS (119 mM NaCl, 4.7 mM KCl, 21 mM NaHCO₃, 1.17 mM MgSO₄, 1.8 mM CaCl₂, 1.18 mM KH₂PO₄, 5 mM glucose, 0.03 mM EDTA) at an intraluminal pressure of 5 mmHg. Following equilibration for 15 min, intraluminal pressure was increased to 110 mmHg, and vessels were stretched to their approximate in vivo length, after which pressure was reduced back to 5 mmHg for an additional 15 min. Vessel viability was assessed for each preparation by evaluating vasoconstriction in response to high extracellular [K⁺] PSS, made isotonic by adjusting [NaCl] (60 mM KCl, 63.7 mM NaCl). Arteries that showed less than 10% constriction in response to elevated [K⁺] were excluded from further investigation.

Myogenic tone was assessed by raising the intraluminal pressure stepwise from 5 mmHg to 120 mmHg in 20 mmHg increments (5 to 20 mmHg for the first step). The active diameter was obtained by allowing vessels to equilibrate for at least 5 minutes at each pressure or until a steady-state diameter was reached. Following completion of the pressure-response study, intraluminal pressure was lowered to 5 mmHg, and arteries were superfused with Ca²⁺-free PSS supplemented with EGTA (2 mM) and the voltage-dependent Ca²⁺ channel blocker diltiazem (10 μM) to inhibit SMC contraction, after which passive diameter was obtained by repeating the stepwise increase in intraluminal pressure. Myogenic tone at each pressure step was calculated as myogenic tone (%) = [1 – (active lumen diameter/passive lumen diameter)] × 100.

Membrane potential:

To measure SMC membrane potential, cerebral arteries were isolated, cannulated, and viability confirmed as described above. SMC membrane potential was recorded from intact pressurized arteries at intraluminal pressures of 20 mmHg and 80 mmHg. SMCs were impaled through the adventitia with glass intracellular microelectrodes (100-200 MΩ) backfilled with 3M KCl. Membrane potential was recorded using an Electro 705 amplifier (World Precision Instruments, Sarasota, FL, USA). Analog output from the amplifier was recorded using Ionwizard software (version 7.5.1.162; IonOptix). Criteria for acceptance of membrane potential recordings were (i) an abrupt negative deflection of potential as the microelectrode was advanced into a cell, (ii) stable membrane potential for at least 30 seconds, and (iii) an abrupt change in the potential to ~0 mV after the electrode was retracted from the cell.

Electrophysiological recordings:

Enzymatically isolated native SMCs were transferred to a recording chamber (Warner Instruments, Holliston, MA, USA) and allowed to adhere to glass coverslips for 20 minutes at room temperature. Recording electrodes (3–5 MΩ) were pulled on a model P-87 micropipette puller (Sutter Instruments, Novato, CA, USA) and polished with a MF-830 MicroForge (Narishige Scientific Instruments Laboratories, Long Island, NY, USA). Currents were recorded at room temperature using an Axopatch 200B amplifier (Molecular Devices, Sunnyvale, CA, USA) equipped with an Axon CV 203BU headstage and Digidata 1440A digitizer (Molecular Devices) for all patch-clamp electrophysiology experiments. Clampex and Clampfit (version 10.2; Molecular Devices) were used for data acquisition and analysis.

The bathing solution composition for perforated-patch recordings of Kv currents, STOCs, and TICCs was 134 mM NaCl, 6 mM KCl, 1 mM MgCl₂, 2 mM CaCl₂, 10 mM HEPES, and 10 mM glucose (pH 7.4, adjusted with NaOH). The patch pipette solution contained 110 mM K-aspartate, 1 mM MgCl₂, 30 mM KCl, 10 mM NaCl, 10 mM HEPES, and 5 μM EGTA (pH 7.2, adjusted with KOH). Amphotericin B (200 μM) was included in the pipette solution to enable electrical access. Whole-cell K⁺ currents were recorded using a step protocol (−60 to +60 mV in 10 mV, 250 ms steps) from a holding potential of −80 mV. The BK channel blocker paxilline (1 μM) was included in the bath solution when Kv currents were recorded. Whole-cell K⁺ currents were recorded in the absence and presence of K_V channel blocker 4-aminopyridine (5 mM), and the Kv component was determined by subtraction. Summary current-voltage (I-V) plots were generated using values obtained from the last 50 ms of each step. STOCs were recorded from SMCs voltage-clamped over a range of membrane potentials (−60 to 0 mV; 10 mV steps). TICCs were recorded from SMCs voltage-clamped at −70 mV; membrane stretch was delivered by applying negative pressure through the recording electrode using a pressure clamp system (HSPC-1; ALA Scientific Instruments Inc., Farmingdale, NY, USA). TICC activity was calculated as the sum of the open channel probability (NP_o) of multiple open states of 1.75 pA ⁶⁵.

Whole-cell BK currents were recorded in the same bath solution Kv currents, STOCs, and TICCs, but the patch pipette solution contained 140 mM KCl, 1.9 mM MgCl₂, 0.075 mM CaCl₂, 10 mM HEPES, 2 mM Na₂ATP, and 0.1 mM EGTA (pH 7.2, adjusted with KOH). Whole-cell K⁺ currents were recorded using a step protocol (−100 to +100 mV in 20 mV, 500 ms steps) from a holding potential of −80 mV. Whole-cell K⁺ currents were recorded in the absence and presence of the BK channel blocker paxilline (1 μM) and the BK component was determined by subtraction. Summary current-voltage (I-V) plots were generated using values obtained from the last 50 ms of each step.

Whole-cell TRPM4 currents were recorded in a bath solution consisting of 156 mM NaCl, 1.5 mM CaCl₂, 10 mM glucose, 10 mM HEPES, and 10 mM TEACl (pH 7.4, adjusted with NaOH). The patch pipette solution contained 156 mM CsCl, 8 mM NaCl, 1 mM MgCl₂, 10 mM HEPES (pH 7.4, adjusted with NaOH), and a free Ca²⁺ concentration of 200 μM, adjusted using the appropriate amounts of CaCl₂ and EGTA, as determined with Max-Chelator software (WEBMAXC standard). Whole-cell cation currents were evoked by applying 400 ms voltage ramps from −100 to +100 mV from a holding potential of −60 mV. Voltage ramps were repeated every 2 s for 300 s. The selective TRPM4 inhibitor 9-phenanthrol (30 μM) was applied after peak TRPM4 current was recorded (~100 s). Whole-cell TRPM4 current amplitude was expressed as the 9-phenanthrol-sensitive current at +100 mV.

Quantitative droplet digital PCR:

Total RNA was extracted from isolated cerebral arteries by homogenization in TRIzol reagent (Invitrogen, Waltham, MA, USA), followed by purification using a Direct-zol RNA microprep kit (Zymo Research, Irvine, CA, USA) with on-column DNAse treatment. RNA concentration was determined using an RNA 6000 Pico Kit run on a Bioanalyzer 2100 running Agilent 2100 Expert Software (B.02.11; Agilent Technologies, Santa Clara, CA, USA). RNA was converted to cDNA with iScript cDNA Supermix (Bio-Rad, Hercules, CA, USA). Quantitative droplet digital PCR (ddPCR) was performed using QX200 ddPCR EvaGreen Supermix (Bio-Rad), custom-designed primers (Supplementary Table 1), and cDNA templates. Generated droplet emulsions were amplified using a C100 Touch Thermal Cycler (Bio-Rad), and the fluorescence intensities of individual droplets were measured with a QX200 Droplet Reader (Bio-Rad) running QuantaSoft (version 1.7.4; Bio-Rad). Analysis was performed with QuantaSoft Analysis Pro (version 1.0596; Bio-Rad).

Ca²⁺ imaging:

Images were acquired using an iXon 897 EMCCD camera (Andor; 16 × 16 μm pixel size) coupled to a spinning-disk confocal head (CSU-X1; Yokogawa) with a 100× oil-immersion objective (Olympus; NA 1.45) at an acquisition rate of 33 frames per second (fps). Custom software (SparkAn) provided by Dr. Adrian D. Bonev (University of Vermont) was used to analyze the properties of Ca²⁺ sparks and whole-cell RyR and IP₃R Ca²⁺ signals.

To image Ca²⁺ sparks, a suspension of freshly isolated SMCs was placed in a glass-bottom 35 mm dish, and cells were allowed to adhere to the glass coverslip for 20 min at room temperature. SMCs were loaded with the Ca²⁺-sensitive fluorophore Fluo-4-AM (1 μM; Invitrogen) in the dark for 20 minutes at room temperature in Mg²⁺-PSS, washed three times with Ca²⁺-containing PSS, and incubated at room temperature for 20 min in the dark to allow sufficient time for Fluo-4 de-esterification. The threshold for Ca²⁺ spark detection was defined as local increases in fluorescence ≥0.2 ΔF/F₀.

To image RyR and IP₃R Ca²⁺ signals, SMCs were plated and loaded with Fluo-4-AM as stated above, but immediately before imaging, cells were washed three times and imaged in Mg²⁺-PSS to eliminate signals resulting from Ca²⁺ influx. Caffeine (10 mM) or U46619 (100 nM; Enzo Biochem, Farmingdale, NY) was applied to the bath.

Statistical analysis:

Summary data are presented as means ± SEM. Statistical analyses were performed, and graphs were constructed using Prism version 9.3.1 (GraphPad Software, San Diego, CA, USA). The value of n refers to the number of cells for patch-clamp electrophysiology and Ca²⁺ imaging experiments, the number of arteries for pressure myography and membrane potential experiments, and the number of animals for histological analyses, magnetic resonance imaging, and ddPCR experiments. Statistical analyses were performed using unpaired Student’s t-tests and one-way and two-way analysis of variance (ANOVA). A P value < 0.05 was considered statistically significant for all analyses.

Data and materials availability:

All data needed to evaluate the conclusions in the paper are present in the paper or the Supplementary Materials. Processed data are available on DRYAD (https://datadryad.org/stash/dataset/doi:10.5061/dryad.k3j9kd5dz). Col4a1 mutant mice are available from D.B.G. under a material transfer agreement with UCSF.