Intramural hematomas and astrocytic infiltration precede perivascular inflammation in a rat model of repetitive low-level blast injury

Abstract

In modern war theaters, exposures to blast overpressures are one of the most common causes of brain injury. These pervasive events result in acute and chronic cerebrovascular degenerative processes. Using a rat model of blast-induced mild traumatic brain injury, we identified intramural periarterial hematomas as early primary acute lesions induced by blast exposures. These lesions resulted in intravascular cell death, cell layer reorganization, and plasma leakage into the intraperiarterial basal membranes that constitute the intraperiarterial drainage system (IPAD). Plasma metalloproteases, including MMP-9, in the IPAD basal membranes may degrade extracellular matrix components compromising normal cerebral interstitial fluid drainage, arterial structure and function leading to chronic vascular degenerative processes. Related subacute effects of blast exposure included increased MMP-9 expression in perivascular reactive astrocytes and the extension of astrocytic processes through the layers of affected vessels. These results, in combination with normal levels of proinflammatory cytokines and the absence of proinflammatory MHC II-expressing microglia, suggest an astrocytic role in the clearing of intravascular hematomas and provide further mechanistic evidence that blast-induced vascular degenerative processes may precede the onset of neurovascular inflammation.

INTRODUCTION

Blast exposure and vascular degenerative processes

Veterans of modern-day wars often suffer blast-induced traumatic brain injury (TBI).¹ Although many Veterans who suffer blast-related TBIs experience improvement of symptoms, others exhibit chronic postconcussive and mental health-related symptoms including anxiety, impulsivity, insomnia, suicidality, cognitive decline, and depression that often worsen over time and are largely refractory to therapy.² Explosives are the most common lethal weapons used in modern war theaters. The magnitude of damage depends on the initial positive pressure peak of the supersonic blast wave originating from an explosive core, the duration of the overpressure, the medium of the explosion, distance from the incident blast wave, and the degree of area confinement or enclosure.³ Acute and chronic vascular degeneration are established major components of blast injury in humans and in animal models.^4–15 Excluding secondary effects from flash burns and penetrating and blunt-force injuries, blast exposure can result in extensive multiorgan trauma. Blast exposure to the central nervous system (CNS) may result in severe vascular injury including edema, intracranial hemorrhage, and vasospasm along with reduced cerebral perfusion and altered contractile properties of large arteries.^9,16 As in humans, acute high-level blast exposure has a prominent hemorrhagic component; in animals this includes venous hemorrhages.¹⁴ Initial blast-induced vascular damage can occur through direct cranial transmission of blast waves or via thoracoabdominal vascular/hydrodynamic mechanisms, whereby a blast wave striking the body causes indirect CNS injury through what has been referred to as a thoracic effect.^11,14,17 Morphological and functional data indicate that both large and small brain vessels are affected.¹⁴ Acute blast exposure has been associated with reduced cerebral blood flow, increased vascular permeability, blood-brain barrier breakdown, and other vascular injuries, including apoptosis of structural elements, capillary strictures, occlusion, glycocalyx alterations, smooth muscle phenotypic changes, vasospasm, rupture, breakdown of the choroid plexus, reduced dilator responses to decreased intravascular pressure, reduced cerebral perfusion, and increased cerebral vascular resistance.^{6,7,13,15,18–30} The evolution of the blast-induced cognitive and behavioral phenotypes in rats seems to overlap with the development of chronic cerebral vascular degenerative processes.^6,15,21,30 The acute vascular injuries are followed by the development of a secondary pathology characterized by perivascular astrocytic degeneration, luminal collapse, disruption of neurovascular interactions, extracellular matrix (ECM) alterations, double-barreled vessels, intraluminal astrocytic processes, vascular smooth muscle degeneration, vascular occlusion by CD34-expressing progenitor cells, vasoconstriction, generalized vascular attenuation, enlarged perivascular spaces, arteriovenous malformations, aneurysms, vascular leakage, perivascular inflammation, and stroke.^21,30 Blast-induced vasospasm has been suggested to initiate a phenotypic switch in vascular smooth muscle cells that causes long-term vascular remodeling.³¹ These combined events can result in disruption not only of cerebral blood circulation but also of the flow of the cerebrospinal fluid (CSF) through the glymphatic system and of the interstitial spinal fluid (ISF) through the intramural periarterial drainage system (IPAD).

CSF and ISF circulation in the brain

Cerebrospinal fluid is an ultrafiltrate of plasma produced by the ependymal cells of the choroid plexus located in the brain lateral ventricles. Cerebrospinal fluid provides the brain mechanical protection and CNS homeostasis as it provides nutrients (glucose, ions), metabolic enzymes (creatine kinase, lactate dehydrogenase), and growth factors (transthyretin, insulin-like growth factor II, and interleukin-1β) to the CNS while clearing waste products.^32–34 Cerebrospinal fluid flows through a convective unidirectional motion from the lateral ventricles through the foramen of Monro, into the third ventricle, passing through the cerebral aqueduct into the fourth ventricle. From the fourth ventricle, it exits through 3 apertures as it enters the cerebral subarachnoid space, the spinal subarachnoid space, and the central canal of the spinal cord.³⁴ Cerebrospinal fluid in the cerebral subarachnoid space flows through the periarterial glymphatic system and into interstitial perivascular spaces along the pial–glial basement membranes that follow the direction of the blood flow. From the periarterial spaces, CSF enters the brain parenchyma where it mixes with the ISF before ultimately exiting through perivenous spaces into the peripheral lymphatic system.³⁵ In turn, drainage of the ISF also occurs through the IPAD system along the basement membranes of capillaries and arteries. The driving force for the IPAD system is provided by the spontaneous contractions of vascular smooth muscle cells.^36,37

The present research is an extension of our previous work on the acute (24 h to 1-week post-exposure) and subacute (6 weeks post-exposure) effects of single and repetitive blast exposures (74.5 kPa) in a male rat model of blast-induced mild TBI (mTBI). Here, we show that blast-induced intramural periarterial hematomas, vascular fragility, blood and plasma leakage into the IPAD, as well as intramural astrocytic vascular penetration are early vascular degenerative processes that precede the onset of perivascular inflammation. As a consequence of blast-induced vascular damage, leaked plasma elements mix within the IPAD compartment and the ISF. This may lead to chronic vascular degeneration as plasma metalloproteases (MMP-3, -8, and -9) and other proteolytic enzymes within the IPAD compartment may degrade structural ECM components, thereby compromising the arterial structure and function, including ISF drainage.

METHODS

Animals

All studies were reviewed and approved by the Institutional Animal Care and Use Committees of the Walter Reed Army Institute of Research/Naval Medical Research Center and the James J. Peters VA Medical Center. Studies were conducted in compliance with the Public Health Service policy on the humane care and use of laboratory animals, the NIH Guide for the Care and Use of Laboratory Animals, and all applicable Federal regulations governing the protection of animals in research. Young adult male Long-Evans hooded rats (250-350 g) of 8 weeks of age were obtained from Charles River Laboratories International (Wilmington, MA, USA) and blast exposed at 10 weeks of age. Animals were housed at a constant 70-72°F temperature on a 12:12 h light cycle. All rats were kept individually in standard clear plastic cages with bedding and nesting paper. Access to food and water was ad libitum.

Blast overpressure exposure

Blast exposures were performed using the shock tube at the Walter Reed Army Institute of Research located at the Naval Medical Research Center (Silver Spring, MD, USA), or a shock tube located at the James J. Peters VA Medical Center (Bronx, NY, USA). Comparable physical characteristics of the blast waves generated by these instruments have been established. Use of the Walter Reed shock tube to deliver blast overpressure injuries to rats has been reported in previous studies.^{15,21,22,30,38–43} Rats were anesthetized with isoflurane and randomly assigned to sham or blast-exposed groups. Animals were placed in a restraint device and secured inside the shock tube in the facing orientation. For blast exposure, the head was facing the source of compressed air without any body shielding, resulting in a full body exposure to the blast wave. Blast-exposed animals received either one 74.5-kPa (10.8 psi) blast exposure (1 × 74.5 kPa) or a total of three 74.5-kPa exposures, with 1 exposure administered daily for 3 consecutive days (3 × 74.5 kPa). Control animals were anesthetized and placed in the blast tube but not subjected to a blast exposure. Groups of control and experimental animals were killed 24 h, 1 week, or 6 weeks post-blast exposures. Animals analyzed in this study 24 h (1 × 74.5 kPa and 3 × 74.5 kPa) and 6 weeks (3 × 74.5 kPa) post-blast were part of larger cohorts previously reported.^{6,15,20–22,30,44}

General histology, stereology, and immunohistochemistry

Coronal sections (50-µm thickness) were prepared with a VT1000S Vibratome (Leica Biosystems). General histological evaluation of hematoxylin & eosin (H&E)-stained sections, sampled every 600 µm and containing the hippocampal region was performed by regular light microscopy. Large vessels with intramural hematomas were scored within the hippocampal stratum lacunosum moleculare (interaural 4.44-6.24 mm). For immunohistochemistry, floating sections were blocked with 10% normal goat serum in 50 mM Tris HCl, pH 7.6, 0.15 M NaCl, 0.3% Triton-X-100 and incubated overnight with the primary antibodies diluted in blocking solution at room temperature. After washing with PBS (6 times for 10 min each), sections were incubated with the appropriate Alexa Fluor (488, 568, and 647)-conjugated secondary antibodies (1:300, ThermoFisher) in blocking solution for 2 h. After washing with PBS (6 times for 10 min each), the sections were mounted with Fluoro-Gel mounting medium (Electron Microscopy Sciences). To visualize nuclei, sections were incubated in 0.1 µg/mL DAPI (4′,6-diamidine-2′-phenylindole dihydrochloride) in PBS in the next-to-last PBS wash. The primary antibodies included a rat monoclonal antiglial fibrillary acidic protein (GFAP, 1:500, clone 2.2B10, gift of Dr Virginia Lee, University of Pennsylvania, Philadelphia PA, USA), rabbit polyclonal anti-GFAP (1:300, G-9269, RRID: AB_477035, Sigma-Aldrich), mouse monoclonal anti-α smooth muscle actin (α-SMA, 1:300, clone 1A4, A-2547, RRID: AB_262054, Sigma-Aldrich), rabbit polyclonal anti-laminin (1:300, L-9393, RRID: AB_477163, Sigma-Aldrich), rabbit polyclonal anti-human fibronectin (1:300, F-3648, RRID: AB_476976, Sigma-Aldrich), mouse monoclonal anti-MMP-9 (1:300, clone 4A3, NBP2-80855, RRID: AB_2811297, Novus), rabbit anti-ionized calcium‐binding adaptor molecule 1 (Iba1, 1:300, 019-19741, RRID: AB_839504, Fujifilm Wako Pure Chemical), mouse monoclonal anti-major histocompatibility complex class II (MHC II, 1:300, clone 3D6, NB200-418, RRID: AB_10001427, Novus), mouse monoclonal anti-phospho-Tau (clone AT270, Thr 181, 1:300, MN1050, RRID: AB_223651, ThermoFisher), rabbit polyclonal anti-collagen type IV (coll IV, 1:300, ab6586, RRID: AB_305584, Abcam). Collagen type IV immunostaining was performed without pepsin pretreatment to visualize ECM remodeling.^45,46 Apoptotic cells were detected with the ApopTag Red In Situ Apoptosis Detection Kit (Millipore). Immunostained sections were imaged with a laser scanning confocal microscope Zeiss LSM 980 with Airyscan 2 (Carl Zeiss Microscopy).

Electron microscopy

Electron microscopy was performed using our protocols optimized to study the ultrastructure of the vasculature.²² Anesthetized rats were transcardially perfused with ice-cold 2.0% glutaraldehyde, 2% paraformaldehyde, 0.1 M sodium phosphate (pH 7.2). Brains were removed, postfixed in the same fixative as above and stored at 4 °C until further processing. Fixed brains were placed on a rat brain slicer matrix, and coronal slices containing the frontal cortex were excised. Sections were washed in 0.1 M sodium cacodylate buffer, pH 7.2 and postfixed with 1% osmium tetroxide in 0.1 M sodium cacodylate buffer, pH 7.2. Sections were washed again in cacodylate buffer, dehydrated through graded ethanol (70%-100%) and propylene oxide series, and resin-infiltrated with Epon (Electron Microscopy Sciences). Material was polymerized in a vacuum oven at 60 °C for 48 h. Semithin (1 µm) toluidine blue-stained sections were used to identify regions of interest. Ultrathin sections (80 nm) were cut with a diamond knife on a Leica UCT ultramicrotome and mounted on copper grids using a Coat-Quick adhesive pen (Electron Microscopy Sciences). Sections were counterstained with uranyl acetate and lead citrate. Frontal cortical sections were imaged on a Hitachi 7700 electron microscope (Hitachi, Ltd) and photographed with an Advantage CCD camera (Advanced Microscopy Techniques). Image brightness and contrast were adjusted using Adobe Photoshop 2022 software (version 23.4.1; Adobe).

RESULTS

Acute brain intraperivascular blood leakage in blast-exposed rats

Light microscopy examination of H&E-stained sections from acutely lesioned animals (24 h to 1-week postblast) revealed the presence of intramural perivascular red blood cells (RBCs) in large vessels of the hippocampal stratum lacunosum moleculare (Figure 1A–H). As expected, these lesions were more frequent in rats exposed to 3 blast exposures (3 × 74.5 kPa) as compared to those subjected to a single (1 × 74.5 kPa) exposure (Figure 2). Moreover, the observed distribution of large hippocampal vessels harboring intramural perivascular blood in 3 × 74.5 kPa exposed animals ranged from 0% to 43%, with no affected vessels found in acute control sham animals (Figure 2). Mainly arteries were affected, but rare affected vessels with thinner medial muscular layers, resembling veins (<1%), could be recognized.

Figure 1.

Identification of blast-induced intramural hematomas in hematoxylin and eosin-stained sections of blast exposed rats. (A) Low magnification (4 × ) of the hippocampus of a blast-exposed rat showing acute vascular intramural hematomas (3 × 74.5 kPa, 1 week post-blast exposure); (B) acute, 1 × 74.5 kPa, 24 h post-exposure, lacunosum moleculare; (C) acute, 1 × 74.5 kPa, 24 h post-exposure, corpus callosum; (D) acute 3 × 74.5 kPa, 24 h post-exposure, lacunosum moleculare; (E–H) acute, 3 × 74.5 kPa, 1 week post-exposure, lacunosum moleculare; (I) subacute, 3 × 74.5 kPa, 6 weeks post-exposure, thalamus; (J–L) sham controls, lacunosum moleculare. Arrows indicate intramural hematomas. Scale bars: A, 0.7 mm; B, 50 µm; C, 50 µm; D, 100 µm; E, 100 µm; F, 50 µm; G, 50 µm; H, 100 µm; I, 25 µm; J, K, 100 µm; L, 400 µm.

Figure 2.

Quantification of intramural hematomas in the hippocampal stratum lacunosum moleculare of acutely and subacutely blast-injured rats. Serial coronal brain sections of blast-exposed rat cohorts were stained with hematoxylin–eosin and the number of arterial vessels with intramural hematomas were determined in the hippocampal stratum lacunosum moleculare (coordinates interaural 4.44-6.24 mm) as described in “Methods”. The graphs show relative percentages of large vessels affected by intramural hematomas in animals exposed acutely (1 × 74.5 kPa, 24 h post-exposure and 3 × 74.5 kPa, 24 h to 1-week post-exposure) and subacutely (3 × 74.5 kPa, 6 weeks post-blast).

Intramural hematomas are associated with blood leakage into the periarterial drainage spaces in the brain of blast-exposed rats

Immunohistochemical analyses with antibodies against rat IgG, α-smooth muscle actin (α-SMA), and the ECM protein collagen type IV or GFAP confirmed the blast-induced intraperivascular leakage of blood elements and its location within the IPAD (Figure 3). Blood elements were mainly contained between the medial α-SMA-immunoreactive smooth muscle layer and the adventitial collagen type IV basement membrane, indicating that the integrity of the medial and intimal layers were compromised by the blast exposures (Figures 3 and 4). These observations were confirmed using the above antibodies in combination with those against the ECM proteins fibronectin and laminin (Figure 5). In a blast-exposed animal examined 1-week post-blast, TUNEL-positive apoptotic nuclei were observed in the intima of endothelial cells and in subadventitial arterial spaces (Figure 6). Transmission electron microscopic observations of the somatosensory/motor cortical region of an acutely blast-injured rat (3 × 74.5 kPa) revealed the presence of a dark amorphous substance in the intraperiarteriolar space above the muscular layer which could correspond to leaked blood elements with cellular debris containing interspersed cellular organelles derived from the affected vascular cellular layers (Figure 7).

Figure 3.

Intraperiarterial plasma leakage in an acutely blast-injured rat. Brain sections of blast-exposed (3 × 74.5 kPa, 1-week post-exposure) and control rats were stained with antibodies against α-smooth muscle actin (green), rat IgG (red) in combination with collagen type IV (A–F) or GFAP (G–I, magenta). DAPI (blue). Confocal micrographs show affected arteries in the hippocampal stratum lacunosum moleculare of the acutely injured rat (A–E and G–H), in comparison to normal arteries from a control rat (F and I). Arrows in C, E, and H show remaining swollen extramural cells surrounded by infiltrating plasma. Scale bar: 10 µm.

Figure 4.

Acute drainage of plasma IgG through the adventitia in an affected artery of a blast-exposed rat. Hippocampal artery of a blast-exposed rat (3 × 74.5 kPa, 1-week post-blast) showing plasma IgG embedded within the arterial adventitia. IgG, red; α-smooth muscle actin, green; GFAP, magenta; DAPI, blue. (A) α-SMA + IgG + GFAP; (B) α-SMA + IgG. Note that the hematoma plasma IgG is contained within the adventitia and does not expand into the neuropil. Scale bar: 10 µm.

Figure 5.

Drainage of plasma IgG through the extracellular matrix of adventitial and medial basement membranes in a hippocampal artery of a blast-exposed rat. Coronal brain sections of a blast-exposed rat (3 × 74.5 kPa, 1-week post-blast) were stained with antibodies against α-smooth muscle actin (green) and IgG (red) in combination with antibodies against ECM components (magenta): laminin (A–C), collagen type IV (D–F), and fibronectin (G–I). Affected arteries with intramural hematomas are shown within the hippocampal stratum lacunosum moleculare indicating colocalization of intramural plasma IgG with ECM basement membrane components. Note the diffuse fibronectin staining within the intramural hematoma suggesting its preferential degradation. Scale bar: 10 µm.

Figure 6.

Apoptotic staining of vascular endothelial cells and perivascular nuclei. Coronal sections from blast-exposed (3 × 74.5 kPa, 1-week post-blast) and control rats were stained for TUNEL (apoptotic cells, red); α-smooth muscle actin (vascular smooth muscle, green); GFAP (astrocytes, magenta) and DAPI (nuclei, blue). Blast-affected vessels in the hippocampus stratum laconosum moleculare are shown. (A) TUNEL (red); (B) α-smooth muscle actin (green); (C) GFAP (magenta); (D) merged. Arrows in A and D identify apoptotic vascular endothelial cells. Arrowheads show perivascular TUNEL-stained nuclei. Scale bar: 10 µm.

Figure 7.

Electron microscopy of an arteriole within the somatosensory cortical region of a blast-exposed rat. Arrows show an electron-dense amorphous substance with interspersed cellular debris external to the smooth muscular layer. Scale bar: 5 µm.

Perivascular hematomas in subacutely injured rats

In rats 6 weeks post-blast (3 × 74.5 kPa) exposure, visible intramural hematomas in the hippocampal stratum lacunosum moleculare were also present (Figure 2). Remnants of intramural blood could also be observed in other brain regions including the thalamus (Figure 1I). Vascular intramural plasma IgG was confined to the adventitial-smooth muscle and endothelial-smooth muscle interfaces (Figures 8 and 9). Rare, affected veins with sub- and extra-adventitial plasma IgG could be identified (Figure 9D). In addition, although less frequent, foci of RBCs could be seen within the periarterial spaces lining the medial and adventitial layers (Figure 10). The presence of morphologically intact RBCs within the periarterial spaces at 6 weeks post-blast reflects either the stability of rat RBCs (half-life ∼60 days⁴⁷) or the occurrence of a more recent hemorrhage due to previous blast-induced vascular fragility.

Figure 8.

Intramural periarterial plasma at 6 weeks post-blast exposure. 3D confocal reconstructions of hippocampal arteries from blast-exposed (A and C) and control (B and D) rats. α-smooth muscle actin, green; collagen type IV, magenta; IgG red; DAPI, blue. Arrows in C show plasma IgG associated with the arterial smooth muscle layer in the blast-exposed animal. Scale bar: 10 µm.

Figure 9.

Intra-arterial and intraperivenous plasma at 6 weeks post-blast exposure. 3D confocal reconstructions illustrating hippocampal intramural hematomas in arteries (A–C) and a rare vein (D). α-smooth muscle actin, green; GFAP, magenta; IgG, red; DAPI, blue. (A) Plasma IgG in between the endothelial and smooth muscle layers; (B) plasma IgG on the external side of the media; (C) penetrating astrocytic processes extending into the intraperiarterial leaked plasma; (D) plasma IgG on both sides of the venous media. Scale bar: 10 µm.

Figure 10.

Intramural hematoma in the hippocampal stratum lacunosum moleculare of a subacutely blast-injured rat. Brain section of a subacutely injured rat (3 × 74.5 kPa, 6 weeks post-exposure) stained with antibodies against phospho-tau (AT270, green), anti-rat IgG (red), and DAPI (blue). (A) Confocal microscopy 3D reconstruction; (B) confocal optical section (0.15 µm thick). Erythrocytes in the intramural hematoma are easily identified by their emitted autofluorescence (arrows). Intramural vascular IgG is indicated (red, arrowheads). Scale bar: 10 µm.

Plasma MMP-9 leakage into the intramural perivascular space

We have previously reported that breakdown of the adventitial collagen type IV ECM is associated with the evolution of blast-induced vascular degenerative processes. As metalloproteases (ie, MMP-3, -8, and -9) are normal components of plasma, we investigated whether plasma MMP-9 and IgG could be found within the perivascular spaces in affected cerebral vasculature of blast-exposed animals by immunohistochemistry. As expected, the presence of leaked intramural plasma MMP-9 colocalized with the IgG in intramural hematomas (Figure 11). Regions neighboring the initial intramural hematoma show that both plasma IgG and MMP-9 were concentrated mainly in the adventitial ECM and basal membranes of mural cells. This indicates that the intramural leaked plasma proteins, IgG and MMP-9, could be draining through the IPAD system. Perivascular plasma leakage seemed to affect arterial endothelial and smooth muscle cells as they frequently lost their elongated fusiform nuclear morphology (Figure 12).

Figure 11.

IgG and MMP-9 in intraperiarterial and adventitial basal membranes. Hippocampal artery of a blast-exposed rat (3 × 74.5 kPa, 1-week postblast) stained with antibodies against IgG (red), MMP-9 (green), and GFAP (magenta). DAPI (blue). Arrows indicate intraperiarterial and adventitial basement membranes. Panel D shows double-staining for IgG and MMP-9. Small arrowheads in B and D indicate RBC’s autofluorescence from the intraperiarterial hematoma. Scale bar: 10 µm.

Figure 12.

Plasma leakage affects the morphology of mural and endothelial cells. (A–D) Periarterial adventitial region of a hippocampal large vessel of an acutely blast-injured rat with plasma leakage (IgG, red) shows different nuclear morphologies (DAPI, blue) of mural and endothelial cells (closed arrows) as compared to the neighboring opposite region without plasma presence (open arrows). IgG, red; DAPI, blue; α-smooth muscle actin, green; GFAP, magenta. Note the round nuclei in the medial/adventitial region with plasma infiltration compared to the normal elongated nuclei of endothelial and smooth muscle cells in the opposite vascular region. Scale bar: 10 µm.

Astrocytic processes in the vascular lumen are associated with intramural vascular blood

As previously reported,^21,22 at 6 weeks post-blast (3 × 74.5 kPa), we observed astrocytic processes extending into the arterial vascular layers and into the lumen of large vessels in the hippocampal lacunosum moleculare. These astrocytic processes extended through the adventitial, vascular smooth muscle, and endothelial layers into the lumen (Figures 13 and 14). In some cases, the affected arterial regions remaining contained intraperivascular erythrocytes without the presence of detectable plasma IgG. This indicates that by 6 weeks post-blast, the plasma IgG in some affected vessels had been reabsorbed or drained through the IPAD system.

Figure 13.

Intramural infiltration of astrocytic processes in a vessel is associated with intramural plasma leakage. (A–D) 3D reconstruction of astrocytic processes extending into the intravascular spaces of a degenerating hippocampal vessel with an intramural hematoma in a blast-exposed rat (3 × 74.5 kPa, 6 weeks post-blast). GFAP, magenta; IgG, red; α-smooth muscle actin, green; DAPI, blue. (A) GFAP + IgG + DAPI; (B) GFAP + DAPI; (C) GFAP  +  αSMA+ IgG + DAPI; (D) GFAP. Arrows indicate astrocytic processes penetrating the periarterial hematoma and smooth muscle arterial layers. Insert in D (confocal optical section, 0.15 µm thick) shows astrocytic processes (GFAP, magenta) penetrating the vascular smooth muscle layer (green). Scale bars: A–D; insert in D, 7.5 µm.

Figure 14.

Intraluminal astrocytic process extensions are subacute effects associated with blast-induced intramural hematomas. Coronal section from a blast-exposed rat (3 × 74.5 kPa, 6 weeks post-blast) was stained with antibodies against GFAP (astrocytes, red) and phosphorylated Tau protein (AT270, Thr181, green). DAPI (nuclei, blue). (A) GFAP; (B) DAPI; (C) GFAP + DAPI; (D) merged: GFAP+DAPI+AT270. Arrows in A, C, and D show intraluminal astrocytic processes; arrowheads in B and D indicate extraluminal autofluorescent erythrocytes. Scale bar: 10 µm.

Subacute increased expression of MMP-9 in perivascular astrocytes of blast-exposed rats

Immunohistochemical observations in the hippocampal lacunosum moleculare of blast-exposed rats (3 × 74.5 kPa, 6 weeks) showed increased expression of MMP-9 in perivascular astrocytes in some vessels (Figure 15). MMP-9 was contained within intracellular and extracellular vesicles. It is known that levels of MMP-9 in reactive astrocytes are increased as a consequence of brain injury, ischemia, or other neurodegenerative conditions.^48,49 Astrocytic MMP-9 compromises the blood-brain barrier and exacerbates intracerebral hemorrhage in animal models.⁵⁰ We also confirmed the absence of proinflammatory MHC II-expressing microglia (MHC II⁺) in the hippocampus of blast-exposed animals (3 × 74.5 kPa, 6 weeks post-blast) from different cohorts (Gama Sosa et al.²⁰; Figure 16).

Figure 15.

Subacute expression of MMP-9 in hippocampal perivascular astrocytes of blast-exposed rats. Sections from blast-exposed (3 × 74.5 kPa, 6 weeks post-blast, panels A–C) and control (panels D–Ff) rats were stained with antibodies against MMP-9 (green) and GFAP (magenta). DAPI (blue). C and F, merged images. Arrows in A–C denote the subacute overexpression of MMP-9 containing vesicles in perivascular astrocytes in the stratum lacunosum moleculare of a blast-exposed rat. Scale bar: 10 µm.

Figure 16.

Lack of activated MHC II-expressing microglia in the hippocampus of subacutely blast-exposed rats. Immunohistochemical analyses of a coronal section from a blast-exposed rat (3 × 74.5 kPa, 6 weeks post-exposure) showed the total absence of activated MHC II⁺ microglia in the hippocampus as previously reported.²⁰ Iba1 (microglia, magenta); MHC II (activated microglia, green); GFAP (astrocytes, red); DAPI (nuclei, blue). (A) 3D confocal reconstruction of an artery in the stratum lacunosum moleculare illustrating the absence of activated proinflammatory MHC II-expressing microglia; (B) 3D confocal reconstruction of a cortical meningeal region harboring MHC II-expressing microglia (arrow, positive control); insert shows optical section (0.15 µm thick) of the meningeal area in B. Scale bar: 10 µm.

DISCUSSION

Intramural hemorrhages

Here we report a type of blast-induced vascular lesion that is probably best described as an intramural hemorrhage or hematoma. These lesions in some ways resemble arterial dissections. Typical arterial dissections occur when tearing of the endothelial intimal layer leads to blood leakage that dissects the endothelial intima and muscular medial layers to form an intramural hematoma that may extend into perivascular spaces.⁶ Arterial dissections result in an altered structure in the endothelial and vascular smooth muscle cells, altered cytoskeletal function, disruption of ECM layers, and dysfunction of the interlinked TGF-β signaling pathways^51,52; all of these features are found after blast injury. Arterial dissections as recognized in human clinical and pathological material, however, typically involve elastic arteries with large muscular layers such as the aorta, carotid, or vertebral arteries and not the smaller arteries and arterioles mostly affected by blast. It is also unclear what the inciting event leading to these lesions after blast exposure might be. Chronic hypertension is a predisposing risk factor for nontraumatic arterial dissections but the vascular abnormalities in these animals lack the typical pathologic changes seen in malignant hypertension, which are predominantly fibrinoid necrosis of vessel walls.⁵³

Abrupt, transient, and severe increases in blood pressure have also been linked to blast exposure. Following blast, an indirect mechanism of injury has been proposed, whereby an overpressure blast wave interacts with the body, compresses the abdomen and chest, and transfers its kinetic energy to the body organs, including the brain and the blood as a fluid medium.⁵⁴ This causes a sudden and abundant stream of blood that rapidly propagates from the torso through the neck into the cerebral vasculature. This blood surge causes a considerable increase in the shear pressure/stress of the arterial wall through the brain vascular network, which may lead to structural alterations in the cerebral arteries and veins.⁵⁵ Whether a transient increase in blood pressure induced by a blood surge might lead to an initial tear and the nidus for dissection-like lesions remains speculative. It is also possible that vascular injury occurs as the blast overpressure wave itself is transmitted through brain. Direct tissue damage from blast waves is most apparent at tissue/fluid interfaces, as would be found at the boundaries of blood vessels. Whatever its mechanism, the abundance of arterial versus vein-like dissections suggest that arterial vessels are structurally more susceptible to the blast overpressure injury.

Blast-induced hippocampal intramural hematomas

Intramural hematomas were commonly detected in the hippocampal stratum lacunosum moleculare where 0%-43% of the arterial large vessels were found to be affected acutely in blast-exposed animals (24 h to 1-week post-exposure). The broad range of affected vessels within groups of blast-exposed animals could be due to the known wide variation in the vascular network architecture, including the arterial circle of Willis anatomy in rats.⁵⁶ Hippocampal blood in the rat is supplied via internal transverse hippocampal arteries that arise from the longitudinal hippocampal artery, itself a branch of the posterior cerebral artery.⁵⁷ The longitudinal hippocampal artery and its branches penetrate the stratum lacunosum moleculare, stratum radiatum and lucidum, and stratum pyramidale. This indicates that in several blast-exposed animals the arterial network derived from the longitudinal hippocampal artery may have been affected by a blood surge through the posterior cerebral artery. In the rat, a blast exposure to the torso increases the intrathoracic pressure, which in turn increases the blood flow pressure in the internal carotid arteries. However, about 46% of the pressure pulse dampens as it travels through the carotids into the basilar artery at the base of the brain. As a result, the basilar artery experiences an increase in blood flow; however, it is not as substantial as the one in the carotids.⁵⁵ The choroid plexus receives its vascular supply from a variety of arteries due to its various locations.⁵⁸ The anterior choroidal artery, which is the most distal branch of the internal carotid artery, supplies the choroid plexus in the lateral ventricles. The posterior choroidal artery, which branches from the posterior cerebral artery, supplies the choroid plexus primarily in the third ventricle but also has a minor role in supplying the choroid plexus in the lateral ventricles. Lastly, the anterior and posterior inferior cerebellar arteries, which branch from the basilar and vertebral arteries respectively, supply the choroid plexus found in the fourth ventricle.⁵⁸ The frequent hemorrhages seen in the dorsal third ventricle in blast-exposed animals could also be related to the blast-induced blood pressure increase through the internal carotids.

Intramural hematomas were also found in other brain regions in blast-exposed animals, including those associated with the corpus callosum, white matter, and the thalamus. The blood supply of the corpus callosum in rodents is provided mainly by the carotid system via the azygos pericallosal artery portion of the anterior cerebral artery distal to the anterior communicating artery.⁵⁹ On the other hand, the basilar artery, posterior cerebral artery, and posterior communicating artery are the major blood supply for the thalamus.⁶⁰ A higher incidence of lesioned vessels was observed in animals that were exposed to 3 blast exposures (3 × 74.5 kPa) as compared to those with a single blast exposure (1 × 74.5 kPa). After an initial blast-induced vascular tear leading to endothelial injury and intravascular blood leakage, later blast exposures could expand the initial injury with subsequent extension of the intramural hemorrhage or create new lesions. As expected, detection of intramural hematomas at 6 weeks post-blast was less compared to those following acute injuries, indicating that these lesions apparently seal themselves and heal over time.

We have previously reported the extension of astrocytic processes into the lumen of blast-affected vasculature.^21,22 In this study, we were able to identify extensions of astrocytic processes through intraperivascular subadventitial plasma, muscle medial and endothelial intimal layers and into the lumen. These abnormalities could be associated with the presence of remaining intraperivascular erythrocytes in the absence of plasma IgG, indicative of IgG reabsorption or drainage as part of the healing process.

Blast-induced intramural hematomas result in alterations of the local intramural structure and drainage of plasma proteins through the IPAD

Our results show that in the rat, blast exposures induce arterial intramural hematomas, primarily from a tear of the intima with subsequent extension of the intramural hemorrhage into the subadventitial plane, resembling a dissecting arterial aneurysm. The initial injury and the subsequent local hematoma result in local cell death with the respective cellular compartments filled with plasma and RBCs close to the initial injury. More distally, cellular compartments and expanded subadventitial spaces become filled with plasma. TUNEL staining confirmed our previous results showing that endothelial and extramural cells are affected by blast exposures.⁶ Immunohistochemical observations show that the intramural plasma IgG clears along the subadventitial and intramural basement membranes of arteries and arterioles, independently of RBCs.

Leaked intramural plasma MMP-9 clears through the IPAD

In addition to IgG, plasma metalloproteinase MMP-9 was shown to clear through the adventitial and intramural basement membranes as these proteins colocalized with the ECM structural proteins collagen type IV, laminin, and fibronectin.^61,62 It seems that the IgG and MMP-9 entered the arterial ECM and were cleared along the arterial and arteriolar basement membranes as both proteins colocalized. The zinc metalloendopeptidase MMP-9 is a crucial mediator of ECM remodeling. Activated MMP-9 has strong proteolytic activity against major components of the vascular ECM including collagens, laminin, and fibronectin. Leaked plasma MMP-9 within the IPAD may result in degradation and remodeling of the arterial ECM, structural disruptions in the intravascular cell layers, and alterations in the normal flow of intramural ISF components leading to the chronic vascular degenerative processes observed.

Structural vascular alterations precede brain inflammation

We hereby confirm that blast-induced vascular structural alterations are well established by 6 weeks post-blast exposures (3 × 74.5 kPa). These structural alterations are sometimes clearly associated with plasma leakage into the arterial IPAD. This suggests that degradation and alterations within the mural basement membranes may be in part responsible for the changes in the cellular architecture of the arterial layers. We were also unable to identify activated proinflammatory MHC II⁺ microglia in the brain parenchyma in animals from 2 different cohorts (Gama Sosa et al.²⁰ and this study; Figure 16). This suggests that in most cases the intramural leaked plasma proteins and RBC lysates in the affected arteries are being cleared through the IPAD system and/or reabsorbed locally before an immune response is launched. Normally, the IPAD system is not competent for the transfer of antigen presenting cells or lymphocytes as it constitutes a unique immunologically privileged microenvironment within the brain.^63,64 However, failure to clear the intramural blood components would result in further mural and adventitial basement membrane degradation. Ultimate damage to the integrity of the adventitial scaffold ECM would result in leakage into the brain parenchyma triggering a perivascular inflammatory response.^20,22

In a rat model of subarachnoid hemorrhage, endovascular perforation resulted in a significant impairment in the IPAD system with decreased ISF clearance rate, marked expansion of perivascular spaces, apoptosis of endothelial cells, activation of astrocytes, overexpression of MMP-9, and loss of collagen type IV.⁶⁵ Similarly, in our rat model of blast mTBI, we observe, associated with the intravascular blood infusion apoptosis of endothelial and extramural cells, activation of perivascular astrocytes with expression of MMP-9, leakage of plasma proteins into the IPAD system, degradation of intravascular ECM components, and extension of astrocytic processes through the arterial layers into the lumen.

CONCLUSION

Overwhelming evidence has shown that cerebral vascular degenerative processes arise from blast overpressure exposures. Our results show that low-level blast exposures result in intravascular blood infusion causing intravascular hematomas with significant alterations in the arterial layers and IPAD system, apoptosis of endothelial cells and extramural cells, activation of astrocytes with over-expression of MMP-9, extension of astrocytic processes into the arterial layers, and degradation of ECM structural components. The role of perivascular astrocytes in the clearing of the intramural hematoma debris needs to be further explored.

FUNDING

This work was supported by Department of Veterans Affairs, Veterans Health Administration, Rehabilitation Research and Development Service awards 1I21 RX003459-01 (M.A.G.S.), 1I21 RX002069-01 (M.A.G.S.), 1I21 RX002876-01 (M.A.G.S.), and 1I01RX003846 (G.A.E.); Department of Veterans Affairs Office of Research and Development Medical Research Service awards 1I01BX004067 (G.A.E.) and 1I01BX005882-01 (G.A.E.); Department of Defense work unit number 0000B999.0000.000.A1503 (S.T.A.) and NIA P30 AG066514 (P.R.H., C.W.Z.). M.A.G.S., R.D.G., R.H.L., D.P., G.M.P., C.W.Z., G.E.A., R.A., U.K., and S.T.A. are employees of the U.S. Government. This work was prepared as part of their official duties. Title 17 U.S.C. §105 provides that “Copyright protection under this title is not available for any work of the United States Government.” Title 17 U.S.C. §101 defines a US Government work as a work prepared by a military service member or employee of the US Government as part of that person’s official duties. The views expressed in this article are those of the authors and do not necessarily reflect the official policy or position of the Department of the Navy, Department of Defense, Department of Veterans Affairs, nor the US Government.

CONFLICTS OF INTEREST

The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.

DATA AVAILABILITY

The datasets generated during and/or analyzed during the current study are available from the corresponding author upon reasonable request.

ETHICS APPROVAL AND CONSENT TO PARTICIPATE

All studies were reviewed and approved by the Institutional Animal Care and Use Committees of the Walter Reed Army Institute of Research/Naval Medical Research Center and the James J. Peters VA Medical Center. Studies were conducted in compliance with the Public Health Service policy on the humane care and use of laboratory animals, the NIH Guide for the Care and Use of Laboratory Animals, and all applicable Federal regulations governing the protection of animals in research.