Vascular and Tissue Changes of Magnetic Susceptibility in the Mouse Brain After Transient Cerebral Ischemia
- 1k Downloads
Quantitative susceptibility mapping (QSM) has been recently introduced as a novel MRI post-processing technique of gradient recalled echo (GRE) data. QSM is useful in depicting both brain anatomy and for detecting abnormalities. Its utility in the context of ischemic stroke has, however, not been extensively characterized so far. In this study, we explored the potential of QSM to characterize vascular and tissue changes in the transient middle cerebral artery occlusion (tMCAO) mouse model of cerebral ischemia. We acquired GRE data of mice brains at different time points after tMCAO, from which we computed QSM and MR frequency maps, and compared these maps with diffusion imaging and multi-slice multi-echo imaging data acquired in the same animals. Prominent vessels with increased magnetic susceptibility were visible surrounding the lesion on both frequency and magnetic susceptibility maps at all time points (mostly visible at > 12 h after reperfusion). Immunohistochemistry revealed the presence of compressed capillaries and dilated larger vessels, suggesting that the appearance of prominent vessels after reestablishment of reperfusion may serve compensatory purposes. In addition, on both contrast maps, tissue regions of decreased magnetic susceptibility were observed at 24 and 48 h after reperfusion that were distinctly different from the lesions seen on maps of the apparent diffusion coefficient and T 2 relaxation time constant. Since QSM can be extracted as an add-on from GRE data and thus requires no additional acquisition time in the course of acute stroke MRI examination, it may provide unique and complementary information during the course of acute stroke MRI examinations.
KeywordsQuantitative susceptibility mapping MR frequency Magnetic resonance imaging Mice Middle cerebral artery occlusion Ischemia
Magnetic resonance imaging (MRI) is an important aid for physicians in the diagnosis and management of patients with acute stroke , providing multiple useful contrasts for assessing hemodynamic function as well as extent and severity of brain injury. In case of ischemic stroke, magnetic resonance angiography, for instance, can identify occlusion of a parent artery , whereas perfusion-weighted imaging (PWI) informs about regional disturbances of cerebral blood supply in hyperacute and acute ischemic stroke . Diffusion-weighted imaging (DWI) has been shown to depict the ischemic lesion in the hyperacute, acute, and subacute stage after an ischemic insult [4, 5, 6, 7]. Analyzing T 1 and T 2 relaxation times has also been used to assess ischemic damage [8, 9].
Bulk magnetic susceptibility is a fundamental physical property representing a materials’ tendency to interact with and distort an applied magnetic field. By applying gradient (recalled) echo (GRE) magnetic resonance-based techniques, such as T 2*-weighted imaging [10, 11], phase imaging [12, 13], and susceptibility-weighted imaging (SWI) [14, 15], it is possible to assess qualitatively magnetic susceptibility variations in the brain. Regarding acute stroke MRI, T 2*-weighted imaging and SWI are used to detect cerebral microbleeds and hemorrhages , where SWI is also used to identify areas of hypoperfusion and to detect acute intravascular emboli . Furthermore, asymmetrical veins between ischemic and normal brain tissues have been demonstrated with SWI, which may add information about local oxygen metabolism [17, 18, 19].
More recently, quantitative susceptibility mapping (QSM) has been introduced as a promising post-processing technique based on GRE data. QSM utilizes the small magnetic field variations arising from the underlying tissue magnetic susceptibility distribution to compute quantitative maps. It provides complementary anatomical contrast of the brain [20, 21], supports identification and characterization of brain lesions [22, 23], but also enables quantification of tissue iron content [24, 25], assessment of functional changes , and quantification of contrast agent concentration . Concerning acute stroke QSM has been shown to be able to assess vessel function and oxygen metabolism in patients with acute stroke [18, 28] as well as in animal models of the disease .
In the present study, we investigated the potential of QSM and MR frequency mapping to assess the evolution of vascular and tissue changes in the mouse brain after transient middle cerebral artery occlusion (tMCAO). We acquired high-resolution GRE, DWI, and multi-slice multi-echo imaging data of mice brains at different time points after reperfusion. On the post-processed QSM and MR frequency maps, magnetic susceptibility and frequency were quantified in prominent vessels and brain tissues in both the ischemic and contralateral hemisphere side. We also evaluated the time courses of the occurrence of regional contrast changes on the frequency, magnetic susceptibility, apparent diffusion coefficient (ADC), and T 2 relaxation time constant maps. Immunohistochemical analyses were performed to assess underlying vascular pathology.
All procedures conformed to the national guidelines of the Swiss Federal act on animal protection and were approved by the Cantonal Veterinary Office Zurich (Permit Number: 18-2014 and 49-2011). All procedures fulfilled the ARRIVE guidelines on reporting animal experiments. Animals were housed in a temperature-controlled room in individually ventilated cages, containing up to five animals per cage, under a 12-h dark/light cycle. Paper tissue was given as environmental enrichment. Access to pelleted food (3437PXL15, CARGILL) and water was provided ad libitum.
Study Design and Ischemia Model
Seventeen male C57Bl6/J mice (Janvier, France, weight range 20–25 g, age range 8–10 weeks) were used. Anesthesia was initiated by using 3% isoflurane (Abbott, Cham, Switzerland) in a mixture of O2 (200 ml/min) and air (800 ml/min) and maintained with 1.5–2% isoflurane. Prior to surgery, a local analgesic (lidocaine, 0.5%, 7 mg/kg) was administered subcutaneously. Temperature was controlled during the surgery and kept constant at 36.5 ± 0.5 °C with a feedback-controlled heating pad system. The surgical procedure was carried out as described [30, 31]. The middle cerebral artery was occluded for 1 h. After surgery, buprenorphine was administered as subcutaneous injection every 6–8 h on the day of surgery (Temgesic, 0.1 mg/kg b.w) and supplied thereafter via the drinking water (1 mg/kg) for 36 h.
tMCAO animals were assessed with MRI at 2 h (n = 3), 4 h (n = 4), 6 h (n = 3), 12 h (n = 4), 24 h (n = 3), and 48 h (n = 6) after reperfusion, with the majority of animals being measured at two time points. ADC maps of all investigated mice were inspected. Animals were analyzed when a lesion was present on the ADC maps.
Magnetic Resonance Imaging
MRI measurements were acquired on a Bruker PharmaScan 47/16 (Bruker BioSpin GmbH, Ettlingen, Germany) operating at 4.7 T and equipped with a cryogenic transmit-receive RF coil . During MRI, mice were spontaneously breathing under isoflurane anesthesia (1.5%). Body temperature was monitored with a rectal temperature probe (MLT415, ADInstruments, Spechbach, Germany) and kept at 36 ± 0.5 °C using a warm water circuit integrated into the animal support.
Anatomical reference data acquired in coronal and sagittal orientations served for accurate positioning of the animal’s head. Global first-order shimming followed by fieldmap-based local shimming was performed on the mouse brain using the automated MAPshim routine to reduce field inhomogeneities.
For DWI, a two-dimensional (2D) multi-segment spin echo sequence with echo planar imaging readout (SE-EPI) was used. The scan parameters were field-of-view (FOV) = 17 mm × 14 mm, acquisition matrix = 128 × 128, nominal in-plane voxel size = 133 μm × 109 μm, 12 slices of 1 mm thickness, and an interslice distance = 1.3 mm, number of segments = 4, echo time (TE) = 27.5 ms, and repetition time (TR) = 3000 ms. Diffusion-encoding was applied in x-, y-, and z-direction (gradient pulse duration = 4 ms, gradient pulse separation = 14 ms) with b values of 100, 200, 400, 600, 800, and 1000 s/mm2, respectively. The acquisition time was 3 min and 48 s.
To extract T 2 relaxation time constant of the brain tissue, a 2D Carr-Purcell-Meiboom-Gill multi-slice multi-echo sequence was applied with FOV = 20 mm × 20 mm, acquisition matrix = 100 × 100, nominal in-plane voxel size = 200 μm × 200 μm, 14 echoes with TE1 = 12 ms and an inter-echo time = 12 ms, TR = 2783 ms, and four averages. The acquisition time was 14 min and 6 s.
For frequency mapping and QSM, a 3D multi-echo GRE sequence was applied using a FOV = 25.6 mm × 25.6 mm × 8 mm and an acquisition matrix = 256 × 256 × 80, resulting in an effectively isotropic spatial resolution of 100 μm × 100 μm × 100 μm. Four echoes were recorded (TE1–4 = 4.5/10.5/16.5/22.5 ms) with TR = 100 ms, flip angle = 15°, monopolar echo readout and no averaging. The acquisition time was 25 min and 36 s.
ADC maps were calculated on a pixel-by-pixel basis with linear regression analysis using the model function:
Single-channel GRE magnitude images were combined using the sum-of-squares method , whereas single-channel GRE phase images were combined by taking the argument of the complex summed single-channel images after subtracting the channel-dependent phase offset estimated in the center of the 3D volume of the first echo . Quantitative susceptibility maps were computed based on these combined phase images. To this end, the combined phase images for each echo were unwrapped using a 3D best-path algorithm , divided by (2π · TE) to obtain the Larmor frequency variation in Hz, and then combined across the different TEs in an optimized way that takes into account the local echo time-dependent contrast-to-noise ratio of the Larmor frequency images . Background frequency contributions were eliminated using sophisticated harmonic artifact removal for phase data (SHARP) , with ten different spherical kernels with varying radii ranging from 100 to 1000 μm , and employing a regularization parameter for truncated singular value decomposition of 0.05. Susceptibility mapping was performed based on SHARP-processed frequency images using homogeneity enabled incremental dipole inversion (HEIDI) .
Significant deviations from the signal distribution were identified in the ischemic hemisphere compared to the unaffected, contralateral hemisphere on all contrasts using Paravision (Bruker) and MRIcron. For quantitative evaluation, volumes-of-interest (VOIs) were drawn around three individual vessel structures that appeared prominent on both the ipsilateral side and the contralateral side if visible, and around one lateral ventricle using MRIcron (www.sph.sc.edu/comd/rorden/mricron). In addition, VOIs were drawn around areas of regional contrast change, as well as the remaining ipsi- and contralateral striatum and cortex. The extent of regional contrast changes was calculated edema corrected as described .
After MRI, few mice were used for immunohistochemistry: 12 h (n = 2), 24 h (n = 3), and 48 h (n = 2) after reperfusion. Animals were deeply anesthetized by intraperitoneal injection of ketamine/xylazine/acepromazine maleate (100/20/3 mg/kg body weight) and decapitated. Brains were immediately removed and snap-frozen in 2-methylbutane (Sigma-Aldrich, Switzerland) cooled with dry ice to − 30 °C and stored at − 80 °C until processing. They were then thawed and fixed in 4% paraformaldehyde (PFA) for 48 h, then trimmed (coronal section) and embedded routinely in paraffin wax. Consecutive sections (3–5 μm) were prepared and, after antigen retrieval via incubation in citrate buffer (pH 6.0) for 20 min at 98 °C, incubated, with rabbit anti-mouse collagen IV (Cat # 2150-1470, AbD Serotec, dil 1:200) for 15–18 h at 4 °C. Subsequently, they were incubated with Envision rabbit, Dako.
Statistical analysis was performed using SigmaPlot 12.5 (Systat Software, San Jose, CA). Frequency values and susceptibility differences of vessels were compared with a Mann-Whitney rank sum test, whereas comparisons between different brain regions were performed using an analysis of variance, followed by Holm-Sidak post hoc test for multiple comparisons. Lesion volumes between different contrasts were compared with Student’s t test.
All post-processed images are made available in a data repository ( https://doi.org/10.6084/m9.figshare.5630071.v1).
Prominent Vessels Within the Ischemic Hemisphere on Frequency and Susceptibility Maps
Examination of brain sections after collagen IV staining demonstrated dilation of larger vessels on the ipsilateral compared to the contralateral side (Fig. 3c). In addition, capillaries showed swollen endothelial cells and narrowed vessel lumen (Fig. 3c, zoom in) compared to contralateral capillaries in equivalent locations, for which vascular lumen was not affected (Fig. 3c, d).
Detection of Tissue Changes on Frequency and Susceptibility Maps
Inspection of the background-field corrected frequency maps revealed ipsilateral tissue areas of decreased frequency values at all investigated time points (Figs. 1 and 2, dotted line). The contrast changes were more apparent at 24 and 48 h after reperfusion compared to earlier time points. Similarly, in QSM areas of low magnetic susceptibility, values were observed at 24 and 48 h after reperfusion, while at earlier time points, such areas were only occasionally visible (Figs. 1 and 2). Areas of contrast change confined within the MCA territory; however, the location and extent largely varied within the groups of animals investigated at an individual time point as well as between the different time points.
Extent of Regional Contrast Changes for Different MRI Contrasts
In the current study, high-resolution MR frequency and quantitative susceptibility maps of the mouse brain were generated after 1 h of tMCAO and different time points after reperfusion. On both maps, prominent vessels were seen with increased frequencies and magnetic susceptibilities. In addition, we observed evolutions of regional contrast, which differed in appearance from those seen on ADC and T 2 relaxation time constant maps.
Increased Susceptibility of Prominent Cerebral Vessels
Appearance of prominent vessels in the brain of patients with ischemic stroke on SW images and quantitative susceptibility maps have been described recently [17, 18, 19]. Their occurrence has been attributed to an increase in the oxygen extraction fraction (OEF) and correlation to misery perfusion indicated by PWI . The increased oxygen extraction leads to higher deoxyhemoglobin concentrations in veins, which increase magnetic susceptibility locally . It has therefore been suggested that these prominent vessels demarcate the penumbra that can be salvaged by vessel recanalization and that SWI can be used to predict infarct growth [17, 19].
In the current study, we also demonstrated the occurrence of prominent vessels in the mouse brain following transient ischemia and reperfusion. Vessels were better visible, appeared wider, and had significantly higher MR frequency values and larger differences of magnetic susceptibility relative to CSF on the ipsilateral side compared to vessels in the contralateral unaffected hemisphere (Figs. 1, 2, and 3). Prominent vessels were mainly observed in the surroundings of the core of the lesion. Reperfusion is associated in part by incomplete blood flow of the microvasculature (no-reflow phenomenon). Examination of brain section stained for the basal membrane demonstrated vasodilation of larger vessels in the ischemic ipsilateral hemisphere, particularly at 24 and 48 h after reperfusion, together with swollen capillaries with narrowed lumen. These findings are in agreement with previous studies which described capillary constriction and impaired capillary reflow as a consequence of pericyte contraction [40, 41]. Thus, our observation of prominent vessels around the lesion after restoration of cerebral blood flow is compatible of lower tissue oxygen availability due to compressed capillaries in the ischemic tissue, which is compensated by an increased oxygen extraction in the surrounding area. As an incomplete blood supply is contributing to the propagation of the ischemic lesion, treatments are required that ameliorate pericyte constrictions and capillary compression to fully restore tissue function after reperfusion. The role of prominent vessels after reperfusion need to be further investigated. Thus, prominent vessels are an important indicator of underlying microvascular pathology and QSM and frequency maps might be used as a proxy of underlying microvascular pathology.
Detection of Tissue Changes on Frequency and Susceptibility Maps
We also observed decreased frequency values and smaller susceptibility differences relative to CSF in brain tissue regions, which has so far not been reported in patients with ischemic stroke. One reason for this could be that we performed our study at a field strength of 4.7 T, which is higher than those commonly used in clinical studies (usually 1.5 T). As the effects of magnetic susceptibility are proportional to the applied magnetic field, we are thus more sensitive regarding detectability of differences in magnetic susceptibility between tissues. Furthermore, we used a higher spatial resolution (100 μm) compared to clinical studies (usually > 1 mm), which again helps to delineate small morphological features. The accompanying loss in signal-to-noise ratio was compensated for by using higher field strength and a cryogenic transmitter-receive RF coil . However, also differences in principle function between our animal model and typical clinical scenarios, including patient selection and stroke etiology, type and location may be responsible for the lack of this observation in patients.
The appearance of regions with decreased frequency values and smaller susceptibility differences was more apparent at 24 and 48 h after reperfusion compared to the earlier investigated time points, but was highly variable between individual animals. Moreover, the growth rate and extent of these areas were significantly smaller on frequency and quantitative susceptibility maps compared to ADC and T 2 maps. The underpinnings of these changes on susceptibility and frequency maps are currently unclear. Since cerebral ischemia is followed by microstructural changes at different stages with edema formation, cell death, and tissue destruction and phagocytosis of debris resulting in cavitation with surrounding gliosis [42, 43, 44], this complex chain of processes is difficult to capture with single frequency or susceptibility values. On the other hand, compartmental water shifts due to cytotoxic and vasogenic edema can be detected by DWI and T 2 mapping, respectively [6, 7]. Interestingly, we did not observe a spatial congruence between regions of increased T 2 and reduced ADC values with areas of reduced MR frequency values and differences in magnetic susceptibility. Thus, both contrasts do likely not represent the ischemic lesion as distinct to T 2-weighted imaging and ADC, where a good correlation to the histopathological lesion has been demonstrated . It might be speculated that in the regions, which are at the core of the lesion, oxygen is not extracted , which increases oxygenated hemoglobin concentrations in those areas and produces a diamagnetic shift of susceptibility . Clearly, the relevance of these speculations and the associated cellular underpinnings need to be investigated in further studies.
Taken together, we were able in our study to identify characteristic changes of magnetic susceptibility in the mouse brain after transient ischemia followed by reperfusion. Since QSM can be reconstructed from GRE data, there is no extra penalty in acquisition time to reconstruct such maps. In addition, several approaches for accelerating GRE acquisitions have been proposed for human imaging [48, 49, 50, 51]. Currently, clinical application of QSM is hampered due to its numerical complexity and computational cost associated with the post-processing procedure. More recently, novel algorithms have been proposed, which allow rapid online reconstruction of susceptibility maps directly after data acquisition, enabling instant evaluation by medical personal . Thus, QSM appears promising as a useful post-processing tool to evaluate GRE data for the diagnosis and follow-up of patients with ischemic stroke.
This study was supported by the Swiss National Science Foundation (Grant PZ00P3_136822) and the Hartmann-Müller Foundation.
Compliance with Ethical Standards
Conflict of Interest
The authors declare that they have no conflict of interest.
All applicable international and national guidelines for the care and use of animals were followed.
- 2.Hilger T, Niessen F, Diedenhofen M, Hossmann KA, Hoehn M. Magnetic resonance angiography of thromboembolic stroke in rats: indicator of recanalization probability and tissue survival after recombinant tissue plasminogen activator treatment. J Cereb Blood Flow Metab. 2002;22(6):652–62. https://doi.org/10.1097/00004647-200206000-00003.CrossRefPubMedGoogle Scholar
- 4.Moseley ME, Cohen Y, Mintorovitch J, Chileuitt L, Shimizu H, Kucharczyk J, et al. Early detection of regional cerebral ischemia in cats: comparison of diffusion- and T2-weighted MRI and spectroscopy. Magn Reson Med. 1990;14(2):330–46. https://doi.org/10.1002/mrm.1910140218.CrossRefPubMedGoogle Scholar
- 8.Calamante F, Lythgoe MF, Pell GS, Thomas DL, King MD, Busza AL, et al. Early changes in water diffusion, perfusion, T1, and T2 during focal cerebral ischemia in the rat studied at 8.5 T. Magn Reson Med. 1999;41(3):479–85. https://doi.org/10.1002/(SICI)1522-2594(199903)41:3<479::AID-MRM9>3.0.CO;2-2.CrossRefPubMedGoogle Scholar
- 9.Grohn OHJ, Kettunen MI, Makela HI, Penttonen M, Pitkanen A, Lukkarinen JA, et al. Early detection of irreversible cerebral ischemia in the rat using dispersion of the magnetic resonance imaging relaxation time, T1rho. J Cereb Blood Flow Metab. 2000;20(10):1457–66. https://doi.org/10.1097/00004647-200010000-00007.CrossRefPubMedGoogle Scholar
- 16.Bai Q, Zhao Z, Sui H, Xie X, Chen J, Yang J, et al. Susceptibility-weighted imaging for cerebral microbleed detection in super-acute ischemic stroke patients treated with intravenous thrombolysis. Neurol Res. 2013;35(6):586–93. https://doi.org/10.1179/1743132813Y.0000000179.CrossRefPubMedGoogle Scholar
- 18.Luo Y, Gong Z, Zhou Y, Chang B, Chai C, Liu T, et al. Increased susceptibility of asymmetrically prominent cortical veins correlates with misery perfusion in patients with occlusion of the middle cerebral artery. Eur Radiol. 2017;27(6):2381–90. https://doi.org/10.1007/s00330-016-4593-y.CrossRefPubMedGoogle Scholar
- 21.Deistung A, Schäfer A, Schweser F, Biedermann U, Turner R, Reichenbach JR. Toward in vivo histology: a comparison of quantitative susceptibility mapping (QSM) with magnitude-, phase-, and R2*-imaging at ultra-high magnetic field strength. NeuroImage. 2013;65:299–314. https://doi.org/10.1016/j.neuroimage.2012.09.055.CrossRefPubMedGoogle Scholar
- 23.Klohs J, Deistung A, Schweser F, Grandjean J, Dominietto M, Waschkies C, et al. Detection of cerebral microbleeds with quantitative susceptibility mapping in the ArcAbeta mouse model of cerebral amyloidosis. J Cereb Blood Flow Metab. 2011;31(12):2282–92. https://doi.org/10.1038/jcbfm.2011.118.CrossRefPubMedPubMedCentralGoogle Scholar
- 24.Langkammer C, Schweser F, Krebs N, Deistung A, Goessler W, Scheurer E, et al. Quantitative susceptibility mapping (QSM) as a means to measure brain iron? A post mortem validation study. NeuroImage. 2012;62(3):1593–9. https://doi.org/10.1016/j.neuroimage.2012.05.049.CrossRefPubMedPubMedCentralGoogle Scholar
- 29.Hsieh MC, Tsai CY, Liao MC, Yang JL, Su CH, Chen JH. Quantitative susceptibility mapping-based microscopy of magnetic resonance venography (QSM-mMRV) for in vivo morphologically and functionally assessing cerebromicrovasculature in rat stroke model. PLoS One. 2016;11(3):e0149602. https://doi.org/10.1371/journal.pone.0149602.CrossRefPubMedPubMedCentralGoogle Scholar
- 30.Vaas M, Enzmann G, Perinat T, Siler U, Reichenbach J, Licha K, et al. Non-invasive near-infrared fluorescence imaging of the neutrophil response in a mouse model of transient cerebral ischaemia. J Cereb Blood Flow Metab. 2017;37(8):2833–47. https://doi.org/10.1177/0271678X16676825.CrossRefPubMedGoogle Scholar
- 34.Hammond KE, Lupo JM, Xu D, Metcalf M, Kelley DA, Pelletier D, et al. Development of a robust method for generating 7.0 T multichannel phase images of the brain with application to normal volunteers and patients with neurological diseases. NeuroImage. 2008;39(4):1682–92. https://doi.org/10.1016/j.neuroimage.2007.10.037.CrossRefPubMedGoogle Scholar
- 39.Gerriets T, Stolz E, Walberer M, Müller C, Kluge A, Bachmann A, et al. Noninvasive quantification of brain edema and the space-occupying effect in rat stroke models using magnetic resonance imaging. Stroke. 2004;35(2):566–71. https://doi.org/10.1161/01.STR.0000113692.38574.57.CrossRefPubMedGoogle Scholar
- 40.Yemisci M, Gursoy-Ozdemir Y, Vural A, Can A, Topalkara K, Dalkara T. Pericyte contraction induced by oxidative-nitrative stress impairs capillary reflow despite successful opening of an occluded cerebral artery. Nat Med. 2009;15(9):1031–7. https://doi.org/10.1038/nm.2022.CrossRefPubMedGoogle Scholar
Open Access This article is distributed under the terms of the Creative Commons Attribution 4.0 International License (http://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and reproduction in any medium, provided you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made.