Skip to main content
Advertisement
Browse Subject Areas
?

Click through the PLOS taxonomy to find articles in your field.

For more information about PLOS Subject Areas, click here.

  • Loading metrics

Increased cerebral vascularization and decreased water exchange across the blood-brain barrier in aquaporin-4 knockout mice

  • Yifan Zhang ,

    Roles Conceptualization, Data curation, Formal analysis, Methodology, Project administration, Software, Validation, Writing – original draft, Writing – review & editing

    yxz137@case.edu (YZ); xin.yu@case.edu (XY)

    Affiliations Department of Biomedical Engineering, Case Western Reserve University, Cleveland, OH, United States of America, Department of Radiology, Case Western Reserve University, Cleveland, OH, United States of America

  • Kui Xu,

    Roles Data curation, Methodology, Software, Validation, Writing – review & editing

    Affiliation Department of Physiology and Biophysics, Case Western Reserve University, Cleveland, OH, United States of America

  • Yuchi Liu,

    Roles Conceptualization, Data curation, Methodology, Software, Validation, Writing – review & editing

    Affiliation Department of Biomedical Engineering, Case Western Reserve University, Cleveland, OH, United States of America

  • Bernadette O. Erokwu,

    Roles Data curation

    Affiliation Department of Radiology, Case Western Reserve University, Cleveland, OH, United States of America

  • Pan Zhao,

    Roles Project administration

    Affiliation Department of Physiology and Biophysics, Case Western Reserve University, Cleveland, OH, United States of America

  • Chris A. Flask,

    Roles Conceptualization, Resources, Writing – review & editing

    Affiliations Department of Biomedical Engineering, Case Western Reserve University, Cleveland, OH, United States of America, Department of Radiology, Case Western Reserve University, Cleveland, OH, United States of America, Department of Pediatrics, Case Western Reserve University, Cleveland, OH, United States of America

  • Ciro Ramos-Estebanez,

    Roles Conceptualization, Writing – review & editing

    Affiliation Department of Neurology, Case Western Reserve University, Cleveland, OH, United States of America

  • George W. Farr,

    Roles Conceptualization, Writing – review & editing

    Affiliations Department of Physiology and Biophysics, Case Western Reserve University, Cleveland, OH, United States of America, Aeromics, LLC, Cleveland, OH, United States of America

  • Joseph C. LaManna,

    Roles Conceptualization, Data curation, Formal analysis, Resources, Validation, Writing – review & editing

    Affiliation Department of Physiology and Biophysics, Case Western Reserve University, Cleveland, OH, United States of America

  • Walter F. Boron,

    Roles Conceptualization, Funding acquisition, Resources, Supervision, Writing – review & editing

    Affiliation Department of Physiology and Biophysics, Case Western Reserve University, Cleveland, OH, United States of America

  • Xin Yu

    Roles Conceptualization, Data curation, Formal analysis, Funding acquisition, Methodology, Resources, Validation, Writing – original draft, Writing – review & editing

    yxz137@case.edu (YZ); xin.yu@case.edu (XY)

    Affiliations Department of Biomedical Engineering, Case Western Reserve University, Cleveland, OH, United States of America, Department of Radiology, Case Western Reserve University, Cleveland, OH, United States of America, Department of Physiology and Biophysics, Case Western Reserve University, Cleveland, OH, United States of America

Abstract

Aquaporin-4 (AQP4) plays an important role in regulating water exchange across the blood-brain barrier (BBB) and brain-cerebrospinal fluid interface. Studies on AQP-4 knockout mice (AQP4-KO) have reported considerable protection from brain edema induced by acute water intoxication and ischemic stroke, identifying AQP4 as a potential target for therapeutic interventions. However, the long-term effects of chronic AQP4 suppression are yet to be elucidated. In the current study, we evaluated the physiological and structural changes in adult AQP4-KO mice using magnetic resonance imaging (MRI) and immunohistochemical analysis. Water exchange across BBB was assessed by tracking an intravenous bolus injection of oxygen-17 (17O) water (H217O) using 17O-MRI. Cerebral blood flow (CBF) was quantified using arterial spin-labeling (ASL) MRI. Capillary density was determined by immunohistochemical staining for glucose transporter-1 (GLUT1). Compared to wildtype control mice, AQP4-KO mice showed a significant reduction in peak and steady-state H217O uptake despite unaltered CBF. Interestingly, a 22% increase in cortical capillary density was observed in AQP4-KO mice. These results suggest that increased cerebral vascularization may be an adaptive response to chronic reduction in water exchange across BBB in AQP4-KO mice.

Introduction

Water movement across the blood-brain barrier (BBB) and brain-cerebrospinal fluid interface is essential for volume and osmotic regulation in the brain. Aquaporins (AQP) are membrane proteins that allow bidirectional water movement across the phospholipid bilayer of the plasma membrane. Among them, aquaporin-4 (AQP4) is the most highly expressed in the perivascular and subpial astrocytic endfoot membranes of the brain [1,2]. Initial examinations of the AQP4 knockout (AQP4-KO) mice revealed no overt neurological abnormalities or defects in osmoregulation [3]. Further studies reported significant protection from brain edema induced by acute water intoxication and ischemic stroke [4,5]. A recent study also reported reduced infarct volume, cerebral edema, and BBB disruption in AQP4-KO mice after transient focal cerebral ischemia [6]. These studies suggest that AQP4 can be a potential target for therapeutic interventions.

The effects of AQP4 deletion on cerebral structure and physiology have also been investigated. Yao et al reported an increase in extracellular volume but unaltered tortuosity in AQP4-KO mice [7]. Saadoun et al observed a lack of macromolecule uptake in constitutive AQP4-KO mice, suggesting preserved BBB integrity [8]. Eilert-Olsen and colleagues also reported preserved ultrastructure of capillary endothelial cells, unaltered expression of tight junction proteins, and unaltered vascular permeability to horseradish peroxidase and Evans blue albumin dye [9]. The preserved BBB function was also found in mice with glial-conditional AQP4 deletion [10]. Interestingly, Igarashi et al observed an increase in regional cerebral blood flow (CBF) in response to acute AQP4 inhibition [11]. However, baseline CBF in AQP4-KO mice was found to be similar to that in WT mice [6]. These results suggest that chronic adaption to AQP4 deletion have led to the normalization of CBF. However, the mechanisms leading to normalized CBF remain unclear.

The aim of this study was to evaluate the adaptive response to AQP4 deletion in adult AQP4-KO mice. Three quantitative measurements were performed to compare the physiological and structural differences between the AQP4-KO mice and their age-matched WT controls. First, water exchange across BBB was evaluated by tracking an intravenous bolus injection of oxygen-17 (17O) enriched water (H217O) using magnetic resonance imaging (MRI). Second, cerebral blood flow (CBF) was quantified using arterial-spin-labeling (ASL) MRI. Finally, capillary density was determined by glucose transporter-1 (GLUT1) immunohistochemistry. Our results suggest that increased capillary density may be an adaptive response to chronic reduction in water exchange across BBB in AQP4-KO mice.

Materials and methods

Animal preparation

This work was performed in accordance with the Animal Research: Reporting In Vivo Experiments (ARRIVE) guidelines. The current study and all the procedures involving animal care/handling were approved by the Animal Care and Use Committee (IACUC) at Case Western Reserve University (Protocol #: 2015–0169).

Two to three months old male AQP4-KO mice (n = 8) and the age-matched wildtype (WT) C57/BL6 mice (n = 6) were characterized. Anesthesia was induced by 3% isoflurane mixed with 100% oxygen, and was maintained with 0.5–1% isoflurane mixed with 30% oxygen and 70% nitrogen. A 30G catheter was inserted to the tail vein for H217O injection. For each mouse, 150 μL of 0.9% saline with 12.6% H217O enrichment was injected within 25 seconds. Throughout the imaging experiment, the body temperature was maintained at 36–37°C by blowing warm air into the magnet through a feedback control system (SA Instruments, Stony Brook, NY). Respiration rate was maintained at 60–90 breaths/min by adjusting the isoflurane level. Upon the completion of MRI scan, the mouse was euthanized with an overdose of isoflurane and the brain was harvested for immunocytochemistry analyses.

Imaging methods

MRI studies were performed on a horizontal Biospec 9.4T scanner (Bruker Inc., Billerica, MA, USA). 1H images were acquired using a 3-cm birdcage coil (Bruker Inc., Billerica, MA, USA). Multi-slice, T2-weighted, axial images were acquired using rapid acquisition with refocused echoes (RARE) sequence.[12] Imaging parameters were: TR/TE, 4000/12 ms; RARE factor, 8; number of averages (NAV), 4; matrix size, 128×128; slice thickness, 3 mm; field of view (FOV), 2×2 cm2.

17O-MRI.

Three-dimensional 17O chemical shift imaging (CSI) data were acquired with custom-made, single-loop, 2.5-cm surface coil at a resonance frequency of 54.27 MHz. At baseline, a fully sampled dataset was acquired with 9×9×5 phase-encoding steps covering an FOV of 2×2×1.5 cm3 that gave rise to a voxel size of 15 μL. The free induction decays (FIDs) were acquired with 256 data points at 16.8 μs dwell time, leading to a spectral width of 30 kHz. Other imaging parameters were: TR/TE, 12.5/0.64 ms; NAV, 128. Total scan time was 10.8 minutes. Following the acquisition of baseline 17O CSI dataset, dynamic 17O-CSI data were acquired repeatedly using a keyhole acquisition scheme to improve the temporal resolution. Specifically, 3×3×3 phase-encoding steps in the central k-space were acquired in 337.5 ms with a single average. H217O injection was initiated 2 min after the start of dynamic acquisition and lasted for 20 to 25 seconds. A total of 3300 datasets were acquired that covered 2 minutes before and 16.8 minutes after H217O injection.

Perfusion MRI.

Single-slice axial arterial spin-labeling (ASL) images were acquired with a flow-sensitive alternating inversion recovery (FAIR) preparation sequence followed by a centrically encoded fast imaging in steady precession (FISP) readout [1315]. Specifically, arterial spin labeling was accomplished by slice-selective and non-selective inversion of the magnetization. The inversion pulse used a 3-ms hyperbolic secant adiabatic inversion pulse, and the inversion thickness was set to three times that of the excitation pulse to ensure a uniform inversion over the entire imaging slice. The FISP acquisition was implemented at 1420 ms following the inversion pulse. Imaging parameters were: flip angle, 60°; TR/TE, 2.4/1.2 ms; NAV, 40; matrix size, 128×128; FOV, 3×3 cm2; slice thickness, 1.5 mm. A proton density image (M0) with no inversion preparation was also acquired using the same imaging parameters. A T1 map with the same spatial resolution was acquired with a FISP-based Look-Locker acquisition consisting of a non-selective adiabatic inversion pulse followed by 10 continuous FISP acquisitions with the following parameters: flip angle, 10°; TR/TE, 4.0/2.0 ms; NAV, 40.

Image processing

All image reconstruction and analysis were performed offline using MATLAB (Natick, MA, USA) based software developed in-house. For the analysis of baseline 17O CSI data, the FIDs were applied with 100 Hz line-broadening, zero-padded to 16×16×5, and Fourier transformed to generate the 3D CSI images. For the reconstruction of the dynamic 17O data, composite images were generated by replacing the central 3×3×3 k-space of the fully sampled 3D CSI dataset with the dynamic data. Each dynamic dataset used the average of 16 consecutive acquisitions, yielding a temporal resolution of 5.4 s.

Once image reconstruction was finished, 17O CSI images were overlaid onto the corresponding 1H images (Fig 1A). Imaging voxels that overlapped with the brain tissue and the ventricles were selected for signal analysis. The kinetics of H217O accumulation and washout in each voxel was quantified by calculating the magnitude of the 17O peak in each voxel, normalized by the baseline signal acquired in the initial 2 minutes before H217O injection (Fig 1B and 1C). Peak H217O uptake was identified as the maximal signal intensity following H217O injection. The normalized time course of the H217O signal (SI) from the peak uptake to the end of the measurement was fit to a mono-exponential function described as Where SIPK and SISS are the peak and steady-state H217O signal, respectively, and τ is the time constant of H217O washout.

thumbnail
Fig 1. 17O-MRI of H217O uptake and washout.

(a) Representative 17O-CSI data overlaid on T2-weighted 1H image. (b,c) Representative time courses of 17O signal changes in a CSF voxel (b) and a brain voxel (c), corresponding to the blue and yellow boxes in (a), respectively. Solid dots represent experimental data and red lines the fitted washout curves. Red arrows indicate the start of H217O injection.

https://doi.org/10.1371/journal.pone.0218415.g001

CBF maps were generated from ASL-MRI data. Briefly, the difference image of the slice-selective and non-selective inversion images (ΔM), the proton density image (M0), and the T1 maps as described previously [13,14]. Subsequently, CBF was calculated using the model proposed by Pell et al[16], using an arterial T1 relaxation time of 2.4 s [17], and a water partition coefficient (λ) of 0.9 [18].

Immunohistochemistry

After the MRI scan, brain tissue was harvested and fixed with 4% paraformaldehyde for determination of capillary density by immunohistochemistry [19]. Specifically, the paraffin-embedded tissue was sectioned at 5-μm thickness and stained with goat polyclonal primary antibody against glucose transporter-1 (GLUT1, 1:200; Santa Cruz, CA). The stained slices were imaged on a Nikon E600 Eclipse microscope (Nikon Instruments Inc, Melville, NY) with a 20X objective and an Aquos EXi camera (QImaging, Surrey, BC, CA). Capillary density was calculated by counting GLUT1 positive capillaries per unit area of brain tissue using a computer-aided image analysis software ImageJ (NIH, Bethesda, MD).

Statistical analysis

Acquisition and analyses of the MRI and immunohistochemical data were performed without knowing the genetic background of the animals. Once data analyses were completed, animal identification was obtained, and the mean and standard deviation were calculated for AQP4-KO and WT mice, respectively. Comparison between AQP4-KO and WT mice, as well as between brain tissue and CSF, used two-tailed Student’s t-tests. A p value of less than 0.05 was considered statistically significant.

Power analysis was performed to determine the sampled size for the current study. Based on the observed inter-subject variation, it was estimated that a minimum number of 6 mice in each group was needed to detect a 20% increase in H217O uptake with 80% power and a significance level of 0.05.

Results

Animal characteristics

The age of the mice at the time of MRI scan was 71.0 ± 8.7 days for AQP4-KO mice and 71.2 ± 2.2 days for WT mice. The AQP4-KO mice showed a ~10% increase in body weight as compared to the age-matched WT mice (26.7 ± 2.3 g versus 24.3 ± 1.4 g, p<0.05).

H217O uptake and washout kinetics

Typical time courses of 17O signal changes in AQP4-KO and WT mice are shown in Fig 1B and 1C. For each mouse, a total of 35 to 50 voxels in brain and 4 to 8 voxels in CSF were analyzed. Following the injection of H217O, both AQP4-KO and WT mice showed a rapid increase in the 17O signal that reached peak intensity within 30 seconds. At least 5 data points were collected during this rapid-increase phase. The group-averaged, baseline-normalized time courses of 17O signal in brain tissue and cerebrospinal fluid (CSF) are shown in Fig 2A. A monotonically decreasing curve was observed during the washout phase. The signal reached steady-state toward the end of data acquisition.

thumbnail
Fig 2. Characteristics of dynamic 17O signal.

(a) Time courses of group-averaged, baseline-normalized 17O signal. (b) Peak H217O uptake. (c) Steady-state H217O level. (d) Time constant of H217O washout. *p<0.05 AQP4-KO versus WT.

https://doi.org/10.1371/journal.pone.0218415.g002

Compared to that in WT mice, the peak 17O signal in AQP4-KO mice was 17% lower (p<0.05, Fig 2B), suggesting reduced H217O uptake. AQP4-KO mice also showed a trend of reduced peak 17O signal in CSF. However, the difference did not reach statistical significance. The steady-state 17O signal showed a similar pattern, with a significant 20% reduction in the brain of AQP4-KO mice (p<0.05, Fig 2C) and a trend of reduction (12%) in the CSF of AQP4-KO mice that did not reach statistical significance. CSF also showed a trend of decreased H217O uptake in both WT and AQP4-KO mice comparing to the brain tissue; however, no statistical significance was detected.

The time constant for H217O washout is shown in Fig 2D. There was no statistically significant difference in the time constant of H217O washout between AQP4-KO and WT mice, either for brain tissue or for CSF. In brain tissue, the time constant was 2.76±0.47 min for WT mice and 2.92±0.76 min for AQP4-KO mice. In CSF, the time constant was 2.82±0.48 and 2.94±0.76 min for WT and AQP4-KO mice, respectively.

Cerebral perfusion

Representative CBF maps from WT and AQP4-KO mice are shown in Fig 3A and 3B, with the corresponding T1 maps shown in Fig 3C and 3D. Mean CBF in the entire brain was calculated by excluding the ventricles. Both mean CBF and T1 values were similar between the AQP4-KO and WT mice (Fig 3E and 3F). Mean CBF was 186±33 ml/min/100 g in WT mice and 168±47 ml/min/100 g in AQP4-KO mice.

thumbnail
Fig 3. Cerebral perfusion.

(a,b) Representative CBF maps of WT and AQP4-KO mice, respectively. (c,d) Representative T1 maps of WT and AQP4-KO mice, respectively. (e) Mean cerebral perfusion (excluding ventricles). (f) Mean T1 relaxation time (excluding ventricles).

https://doi.org/10.1371/journal.pone.0218415.g003

Capillary density

Fig 4 shows results of GLUT1 staining and capillary density quantified by counting GLUT1 positive capillaries. Compared to WT mice, AQP4-KO mice showed a significant increase in capillary density by about 22%. The number of GLUT1 positive capillaries were 406±32 and 495 ± 45 /mm2 in WT and AQP4-KO mice, respectively (P<0.05).

thumbnail
Fig 4. Microvascular density in the cerebral cortex in WT and AQP4-KO mice.

(a,b) Representative images of GLUT1 immunohistochemical staining of WT (a) and AQP4-KO (b) mice, respectively. (c) Capillary density (Number/mm2) as quantified by the GLUT1 positive staining in brain cortex. *p<0.05 AQP4-KO versus WT.

https://doi.org/10.1371/journal.pone.0218415.g004

Discussion

In this study, we characterized the sequestration and subsequent washout of H217O in the brain of AQP4-KO mice by directly observing kinetic changes in the 17O signal after an intravenous bolus injection of H217O. By using a keyhole-based data acquisition scheme, a 5.4-s temporal resolution was achieved that enabled the delineation of the rapid signal increase immediately after H217O injection. Consistent with a previous study using T2-weighted proton MRI for indirect H217O detection [20], our results also showed decreased H217O uptake and retention in AQP4-KO mice. This reduction in H217O uptake is a direct consequence of the lack of AQP4-mediated water movement across the BBB. It is also consistent with a previous study that reported reduced BBB water uptake in mice with glial-conditional AQP4 deletion [10].

It is worth noting that CSF also showed a trend of reduced H217O uptake in AQP4-KO mice, although no statistical difference was detected in the current study. Previously, Igarashi et al evaluated H217O uptake in AQP4-KO mice indirectly by T2-weighted proton MRI. They observed decreased H217O uptake into the third ventricle of AQP4-KO mice [20]. A possible explanation for this discrepancy could be the low spatial resolution in our current study. Due to the low sensitivity of 17O-MRI, a keyhole-based method was used to capture the kinetics of 17O signal. Since only the center of the k-space was updated during dynamic data acquisition, the dynamic information is only of low spatial resolution [21], which may limit the ability of this methods to detect changes in small regions such as CSF. Future improvement could be made to improve the spatial resolution by using other fast imaging methods such as ultra-short echo-time (UTE) and compressed sensing.

Despite reduced H217O uptake by AQP4-KO mice, the time constant of H217O washout was similar between AQP4-KO and WT mice. Previous studies using oxygen-15 labeled water (H215O) and positron emission tomography (PET) have demonstrated that the kinetics of H215O is dominated by cerebral perfusion [22,23]. Like H215O, the washout kinetics of H217O is also primarily governed by cerebral perfusion, with only a slight contribution from water movement across the BBB. Thus, the lack of difference in the kinetics of H217O washout suggests that cerebral perfusion was similar in AQP4-KO and WT mice. Indeed, CBF measured by ASL-MRI in the current study was similar between AQP4-KO and WT mice. A recent study employing a multi-echo ASL technique also reported unaltered CBF in AQP4-KO mice, although the exchange time of labeled intravascular water across BBB was significantly increased by 31% [24]. Similar CBF between AQP4-KO and WT mice was also observed by Yao et al using laser Doppler flowmetry [6]. On the other hand, a previous study has reported that acute inhibition of AQP4 by the putative AQP4 inhibitor TGN-020 induced a significant increase in CBF in WT mice [11]. Hence, the lack of change in baseline CBF suggests that chronic AQP4 deletion can lead to an adaptive response toward normalized cerebral perfusion.

A major finding of the current study is the increase in capillary density in AQP4-KO mice assessed by GLUT1 immunocytochemistry. Comparing to other methods of assessing vessel density, e.g. fluorescent microscopy [25], GLUT1 immunocytochemistry is specific to micro vessels with blood-brain barrier but not to arterioles [26]. With unaltered CBF, this increase in capillary density will lead to increased mean transit time (MTT = cerebral blood volume/CBF) in AQP4-KO mice. Hence, it is possible that chronic AQP4 deletion may have led to the adaptive response to compensate for reduced water exchange at BBB by increasing the time for exchange via increased microvasculature. Another hypothesis that is worth investigating is the role of AQP4 in facilitating oxygen diffusion. The central pore in an AQP4 tetramer has been shown to be permeable to gases such as oxygen, carbon dioxide, nitric oxide, and ammonia [2729]. This newly identified role of AQP4 suggests that AQP4 deletion may lead to insufficient oxygen supply when metabolic demand is high. Indeed, Thrane et al have reported impaired oxygenation in areas remote from brain microvessels in AQP4-KO mice during cortical spreading depression, suggesting the involvement of AQP4 in oxygen diffusion in vivo [30]. Hence, the increase in capillaries may also be in response to reduced oxygen diffusion associated with AQP4 deletion. Further investigation of cerebral metabolic activities in AQP4-KO mice will allow new insights into the role of AQP4 in oxygen delivery and regulating cerebral metabolism.

Our current study used H217O injected via tail vein as a tracer. While the results suggest that 17O-MRI is sensitive to altered water exchange across BBB, it should be noted that our current study does not provide quantitative assessment of BBB permeability to water. Tail vein injection causes a dispersion of the arterial input function, which needs to be measured experimentally in order to quantify BBB permeability through kinetic modeling. Due to the low-resolution of 17O-MRI, dynamic changes of 17O signal in carotid artery cannot be reliably measured with the current method. A feasible approach to obtain the arterial input function would be to measure 17O signal in ventricular cavity using a two-coil system and interleaved acquisition. Alternatively, intra-carotid injection allows the tracking of the first pass of tracer, which can provide reliable measurement of the permeability-surface area product, as was demonstrated in PET studies using H215O as a tracer [31]. This approach requires further improvement in temporal resolution. Future development that combines compressed sensing with UTE may enable quantitative assessment of BBB permeability by 17O-MRI [32].

In conclusion, we observed increased capillary densities and reduced water exchange across BBB in AQP4-KO mice. These findings suggest an adaptive response to chronic AQP4 deletion toward increased MTT for increased time of water exchange and perhaps oxygen diffusion as well across BBB. Future work on oxygen delivery and cerebral metabolism in AQP4-KO mice is warranted to elucidate the role of AQP4 as a gas channel in addition to a water channel.

Acknowledgments

The authors thank Dr. Alan Verkman for providing the AQP4-KO mice.

References

  1. 1. Jung JS, Bhat R V, Preston GM, Guggino WB, Baraban JM, Agre P. Molecular characterization of an aquaporin cDNA from brain: candidate osmoreceptor and regulator of water balance. Proc Natl Acad Sci U S A. 1994;91: 13052–6. pmid:7528931
  2. 2. Hasegawa H, Ma T, Skach W, Matthay MA, Verkman AS. Molecular cloning of a mercurial-insensitive water channel expressed in selected water-transporting tissues. J Biol Chem. 1994;269: 5497–500. pmid:7509789
  3. 3. Ma T, Yang B, Gillespie A, Carlson EJ, Epstein CJ, Verkman AS. Generation and phenotype of a transgenic knockout mouse lacking the mercurial-insensitive water channel aquaporin-4. J Clin Invest. 1997;100: 957–62. pmid:9276712
  4. 4. Verkman AS, Yang B, Song Y, Manley GT, Ma T. Role of water channels in fluid transport studied by phenotype analysis of aquaporin knockout mice. Exp Physiol. 2000;85 Spec No: 233S–241S. pmid:10795927
  5. 5. Manley GT, Fujimura M, Ma T, Noshita N, Filiz F, Bollen AW, et al. Aquaporin-4 deletion in mice reduces brain edema after acute water intoxication and ischemic stroke. Nat Med. 2000;6: 159–63. pmid:10655103
  6. 6. Yao X, Derugin N, Manley GT, Verkman AS. Reduced brain edema and infarct volume in aquaporin-4 deficient mice after transient focal cerebral ischemia. Neurosci Lett. 2015;584: 368–72. pmid:25449874
  7. 7. Yao X, Hrabetová S, Nicholson C, Manley GT. Aquaporin-4-deficient mice have increased extracellular space without tortuosity change. J Neurosci. 2008;28: 5460–4. pmid:18495879
  8. 8. Saadoun S, Tait MJ, Reza A, Davies DC, Bell BA, Verkman AS, et al. AQP4 gene deletion in mice does not alter blood-brain barrier integrity or brain morphology. Neuroscience. IBRO; 2009;161: 764–72. pmid:19345723
  9. 9. Eilert-Olsen M, Haj-Yasein NN, Vindedal GF, Enger R, Gundersen GA, Hoddevik EH, et al. Deletion of aquaporin-4 changes the perivascular glial protein scaffold without disrupting the brain endothelial barrier. Glia. 2012;60: 432–40. pmid:22131281
  10. 10. Haj-Yasein NN, Vindedal GF, Eilert-Olsen M, Gundersen GA, Skare Ø, Laake P, et al. Glial-conditional deletion of aquaporin-4 (Aqp4) reduces blood-brain water uptake and confers barrier function on perivascular astrocyte endfeet. Proc Natl Acad Sci U S A. 2011;108: 17815–20. pmid:21990350
  11. 11. Igarashi H, Tsujita M, Suzuki Y, Kwee IL, Nakada T. Inhibition of aquaporin-4 significantly increases regional cerebral blood flow. Neuroreport. 2013;24: 324–8. pmid:23462267
  12. 12. Hennig J, Nauerth A, Friedburg H. RARE imaging: A fast imaging method for clinical MR. Magn Reson Med. 1986;3: 823–833. pmid:3821461
  13. 13. Gao Y, Goodnough CL, Erokwu BO, Farr GW, Darrah R, Lu L, et al. Arterial spin labeling-fast imaging with steady-state free precession (ASL-FISP): a rapid and quantitative perfusion technique for high-field MRI. NMR Biomed. 2014; pmid:24891124
  14. 14. Goodnough CL, Gao Y, Li X, Qutaish MQ, Goodnough LH, Molter J, et al. Lack of dystrophin results in abnormal cerebral diffusion and perfusion in vivo. Neuroimage. The Authors; 2014;102: 809–816. pmid:25213753
  15. 15. Tsekos N V., Zhang F, Merkle H, Nagayama M, Iadecola C, Kim SG. Quantitative measurements of cerebral blood flow in rats using the FAIR technique: correlation with previous iodoantipyrine autoradiographic studies. Magn Reson Med. 1998;39: 564–73. pmid:9543418
  16. 16. Pell GS, Thomas DL, Lythgoe MF, Calamante F, Howseman a M, Gadian DG, et al. Implementation of quantitative FAIR perfusion imaging with a short repetition time in time-course studies. Magn Reson Med. 1999;41: 829–40. pmid:10332861
  17. 17. Dobre MC, Uğurbil K, Marjanska M. Determination of blood longitudinal relaxation time (T1) at high magnetic field strengths. Magn Reson Imaging. 2007;25: 733–5. pmid:17540286
  18. 18. Herscovitch P, Raichle ME. What Is the Correct Value for the Brain Blood Partition-Coefficient for Water. J Cereb Blood Flow Metab. 1985;5: 65–69. pmid:3871783
  19. 19. Tsipis CP, Sun X, Xu K, Lamanna JC. Hypoxia-induced angiogenesis and capillary density determination. Methods Mol Biol. 2014;1135: 69–80. pmid:24510855
  20. 20. Igarashi H, Tsujita M, Kwee IL, Nakada T. Water influx into cerebrospinal fluid is primarily controlled by aquaporin-4, not by aquaporin-1: 17O JJVCPE MRI study in knockout mice. Neuroreport. 2014;25: 39–43. pmid:24231830
  21. 21. Bishop JE, Santyr GE, Kelcz F, Plewes DB. Limitations of the keyhole technique for quantitative dynamic contrast-enhanced breast MRI. J Magn Reson Imaging. 1997;7: 716–23. pmid:9243393
  22. 22. Herscovitch P, Markham J, Raichle ME. Brain blood flow measured with intravenous H2(15)O. I. Theory and error analysis. J Nucl Med. 1983;24: 782–9. pmid:6604139
  23. 23. Raichle ME, Martin WRW, Herscovitch P, Mintun M a, Markham J, Herscovltch P, et al. Brain blood flow measured with intravenous H2(15)O. II. Implementation and validation. J Nucl Med. 1983;24: 790–8. pmid:6604140
  24. 24. Ohene Y, Harrison IF, Nahavandi P, Ismail O, Bird E V., Ottersen OP, et al. Non-invasive MRI of brain clearance pathways using multiple echo time arterial spin labelling: an aquaporin-4 study. Neuroimage. Elsevier Ltd; 2019;188: 515–523. pmid:30557661
  25. 25. Bohn KA, Adkins CE, Mittapalli RK, Terrell-Hall TB, Mohammad AS, Shah N, et al. Semi-automated rapid quantification of brain vessel density utilizing fluorescent microscopy. J Neurosci Methods. Elsevier B.V.; 2016;270: 124–131. pmid:27321229
  26. 26. Patching SG. Glucose Transporters at the Blood-Brain Barrier: Function, Regulation and Gateways for Drug Delivery. Mol Neurobiol. Molecular Neurobiology; 2017;54: 1046–1077. pmid:26801191
  27. 27. Friscourt F, Badaut J. Aquaporins through the brain in health and disease: From water to gas movements. J Neurosci Res. 2018;96: 177–179. pmid:28960397
  28. 28. Rodriguez-Grande B, Konsman JP. Gas Diffusion in the CNS. J Neurosci Res. 2018;96: 207–218. pmid:28504343
  29. 29. Boron WF. Sharpey-Schafer Lecture: Gas channels. Exp Physiol. 2010;95: 1107–1130. pmid:20851859
  30. 30. Thrane AS, Takano T, Thrane VR, Wang F, Peng W, Ottersen OP, et al. In vivo NADH fluorescence imaging indicates effect of aquaporin-4 deletion on oxygen microdistribution in cortical spreading depression. J Cereb Blood Flow Metab. Nature Publishing Group; 2013;33: 996–999. pmid:23611872
  31. 31. Herscovitch P, Raichle ME, Kilbourn MR, Welch MJ. Positron emission tomographic measurement of cerebral blood flow and permeability-surface area product of water using [15O]water and [11C]butanol. J Cereb Blood Flow Metab. 1987;7: 527–542. pmid:3498732
  32. 32. Zhang J, Feng L, Otazo R, Kim SG. Rapid dynamic contrast-enhanced MRI for small animals at 7T using 3D ultra-short echo time and golden-angle radial sparse parallel MRI. Magn Reson Med. 2018; 140–152. pmid:30058079