Mild traumatic brain injury induces pericyte detachment independent of stroke vulnerability

Abstract

Mild traumatic brain injury (mTBI) is an independent risk factor for ischemic stroke and can result in poorer outcomes- an effect presumed to involve the cerebral vasculature. Here we tested the hypothesis that mTBI-induced pericyte detachment from the cerebrovascular endothelium is responsible for worsened stroke outcomes. We performed a mild closed-head injury and/or treated C57/bl6 mice with imatinib mesylate, a tyrosine kinase inhibitor that induces pericyte detachment. The time course of pericyte detachment was assessed 7, 14, and 28 days post injury (DPI). To test the role of pericytes in TBI-induced exacerbation of ischemic stroke outcomes, we induced mTBI and/or treated mice with imatinib for one week prior to transient middle cerebral artery occlusion. We found that injury promoted pericyte detachment from the vasculature commensurate with the levels of detachment seen in imatinib-only treated animals, and that the detachment persisted for at least 14DPI, but recovered to sham levels by 28DPI. Further, mTBI, but not imatinib-induced pericyte detachment, increased infarct volume. Thus, we conclude that the transient detachment of pericytes caused by mTBI may not be sufficient to exacerbate subsequent ischemic stroke damage. These data have important implications for understanding cerebrovascular dysfunction following mTBI and potential mechanisms of increased risk for future ischemic strokes.

Introduction:

Traumatic brain injury (TBI) alters the incidence and functional outcomes of ischemic stroke in experimental animals [1, 2] and in clinical settings [3–6]. Mild TBI can impair cerebrovascular function in a number of ways, including promoting intraluminal protein accumulation, impairing angiogenesis, and damaging the blood-brain barrier (BBB) [2]. TBI both directly damages elements of the cerebrovasculature and sets in motion processes, including inflammation, that can further exacerbate vascular injury [7–10].

The neurovascular unit is comprised of cell types that together regulate cerebral blood flow including neurons, astrocytes, endothelial cells, pericytes and/or smooth muscle cells depending on the size of the vessel [11]. These cells regulate cerebral blood flow (CBF) to ensure that neural tissue receives the metabolic substrates necessary for optimal function [12, 13]. TBI-mediated damage to the cerebral vasculature can result in vascular inflammation, aberrant intravascular coagulation, and mismatches between metabolic demand and blood flow which together can increase the risk of ischemia [14–16].

Pericytes are cells that closely associate with the cerebrovascular endothelium and play important roles in the regulation of microvascular blood flow. Mild or moderate brain injuries can produce pericyte detachment from vascular walls and impair endothelial-pericyte communications [14, 17]. Pericyte detachment and dysfunction could impair BBB integrity and dysregulate cerebral blood flow [9, 17–20]. However, it remains unspecified whether impairments in endothelial-pericyte interactions after TBI contributes to the risk of subsequent ischemic results. Therefore, we opted to investigate the role of pericyte detachment in these experiments in the context of both mild traumatic brain injury (mTBI) and subsequent ischemic stroke.

Here, we tested the hypothesis that pharmacologically or TBI-induced pericyte detachment would exacerbate ischemic cell death and functional deficits. Pharmacological induction of pericyte detachment was achieved using imatinib, a tyrosine kinase inhibitor that targets the PDGFR-ß receptor on pericytes, and induces pericyte detachment in the cerebrovasculature, even when administered systemically [21, 22]. Specifically, young adult mice of both sexes underwent a brain injury and were treated with imatinib for one week following TBI before an undergoing a transient middle cerebral artery occlusion (tMCAO). We then measured ischemic infarcts and motor deficits as well as conducted immunohistochemical analyses for interactions between endothelial cells and pericytes.

Methods:

All procedures were approved by the West Virginia University Institutional Animal Care and Use committee and were conducted in accordance with NIH guidelines. C57Bl/6 mice derived from progenitors acquired from Jackson Laboratory were bred in our colony at WVU. Mice of both sexes were randomly assigned to receive either the TBI or control procedure and divided in half again to receive either imatinib or vehicle injection (Figure 1). This produced four groups of animals per sex SHAM (males n = 10, females n = 9), TBI (males n = 10, females n = 9), Imat, (males n = 6, females n = 6) and TBI-Imat (males n = 6, females n = 7). We included 68 animals overall in this experiment, with 5 animals being euthanized after reaching predefined early removal criteria such as immobility, skull fracture, or excessive weight loss.

Figure 1.

Experimental timeline. Cohort 1: Impacts of mild traumatic brain injury (mTBI) and Imatinib treatments on pericyte detachment from cerebrovasculature. Cohort 2: Timeline of mTBI-induced pericyte detachment and recovery. Cohort 3: Impacts of mTBI and Imatinib on ischemic outcomes.

A second cohort of mice underwent experimental mTBI or an equivalent sham surgery before collecting tissue at different time points after injury (7 days post-injury (DPI), 14 DPI, or 28 DPI). This resulted in 6 total groups per sex: 7DPI-TBI (males n = 6, females = 6), 7 DPI-Sham (males = 6, females = 6), 14DPI-TBI (males n = 5, females n = 5), 14-Sham (males n = 5, females n = 5), 28DPI-TBI (males n = 6, females n = 6), and 28-Sham (males n = 5, females = 6). We included 72 total animals in this study, with 3 animals having tissue excluded for poor perfusions in imaging sessions.

The final cohort of mice underwent the same manipulations as cohort 1 followed by MCAO, resulting in the same four groups of animals per sex SHAM-MCAO (males n = 9, females n = 6), TBI-MCAO (males n = 9, females = 7), Imat-MCAO (males n = 8, females = 8), and TBI-Imat-MCAO (males n = 6, females = 7). We performed surgery on 72 mice for this experiment, and 12 mice died over the course of the study (See Figure 1 for experimental timelines).

Traumatic brain injury

Mice were randomly assigned to an experimental group and then placed into a stereotaxic frame under inhaled isoflurane anesthesia (3% induction, 1.5% maintenance). Skin was retracted to expose the skull prior to the injury. The injury was induced with an Impact One device (Leica Biosystems, Buffalo Grove, IL). A 3 mm diameter impactor tip was retracted and driven into the skull along the midline (=−2 mm AP relative to bregma) to a depth of 1.2 mm at a rate of 3 mm/s (dwell time: 30 msec). Local analgesia was provided with subcutaneous bupivacaine (1.5mg/kg) at the incision site. The skin was sutured with 6/0 nylon suture and mice were returned to their cages. The control procedure was performed identically but the impactor was not driven into the skull (just placed on the surface and retracted).

Middle cerebral artery occlusion

Briefly, following a 7-day recovery period after TBI, mice in cohort 3 (Fig. 1) were reanesthetized with inhaled isoflurane (3% induction, 1.5% maintenance) and a unilateral middle cerebral artery occlusion (MCAO) was achieved by insertion of a 0.23 mm occluder (Doccol) into the right middle cerebral artery via the internal carotid artery and extending 6 mm beyond the internal carotid-pterygopalatine artery bifurcation. The occluder was secured and the mouse was returned to its home cage for 1 h. Following the 1-h occlusion period, the mouse was reanesthetized, and reperfusion initiated by removal of the occluder. Local analgesia was provided with subcutaneous bupivacaine (1.5mg/kg) at the incision site. MCAO mice were included in the study if a measurable infarct was present.

Treatment

Following mTBI or equivalent control procedure, mice were separated into two groups. One group received intraperitoneal injections once per day for one week of imatinib mesylate (Thermo Scientific, Waltham, MA). Imatinib was administered at 75 mg/kg of body weight dissolved in saline [23]. The second group of mice received an equivalent volume of a saline-only vehicle injection.

Rotarod Testing

Motor deficits following mTBI and tMCAO were assessed using the rotarod. Briefly, mice were placed on a rotating rod (Ugo Basile, Gemonio, Italy) that accelerates from 5–80 RPM over 3 minutes from a base rotational speed of 5 RPM to a maximum rotation of 80 RPM. Three trials were conducted per experimental timepoint, and the average latency to fall and the corresponding speed of the fall time across the three trials were recorded. A baseline measurement was conducted on mice prior to any injury or manipulations, with a secondary measurement conducted seven days following mTBI on the morning of tMCAO induction, and a final measurement collected on the day of tissue collection.

Determination of infarct volume

Three days after tMCAO, mice of cohort 3 were euthanized by rapid decapitation; brains were sectioned into 4 2 mm-thick coronal sections through the forebrain and incubated in 1% TTC (2,3,5-triphenyltetrazolium chloride in saline) for 10 min at 37°C. TTC produces a pink formazan product in the presence of live mitochondria, thus ischemic lesions are visualized as unstained white tissue. Slices were post-fixed in 4% paraformaldehyde overnight, photographed, and infarct area was outlined on both sides of each slice using Fiji imaging software [24]. Infarct size was determined after correcting for edema and reported as percent of the contralateral hemisphere using the formula: (1-(total ipsilateral hemisphere – infarct)/total contralateral hemisphere) *100.

Immunohistochemistry

Mice in cohorts 1 and 2 were transcardially perfused with 4% paraformaldehyde and the brains were postfixed, cryoprotected and frozen before being cut on a cryostat at 40μm through the forebrain.

For immunofluorescent staining, slices were washed with 0.1M PBS-T (phosphate buffered saline with 0.01% Triton-X) and blocked in 1% normal horse serum. After blocking, slices were incubated overnight at room temperature in primary antibodies: tomato lectin conjugated to Dylight 594 (TL) (Vector Laboratories, Burlingame, CA, 1:200), which marks blood vessels, and Aminopeptidase N CD13 (R&D Systems, Minneapolis, MN, 1:200), a surface marker that labels brain pericytes. CD13 works as well as several other immunohistochemical antibodies (PDGFRß, NG2, Desmin), and has been shown to be fairly selective for pericytes [25, 26]. The next day, sections were washed with PBS-T, incubated with a fluorescent secondary antibody (donkey anti-goat 488, Life Technologies, Eugene, OR,1:200), mounted, and coverslipped with Fluoromount with DAPI mounting medium (Invitrogen, Carlsbad, CA).

Imaging and Analysis

All immunohistochemistry (between bregma −1.5mm and −2.0mm) was imaged and stitched on an Olympus VS-120 microscope with a 10× Plan S Apo/0.40 NA objective using a Monochrome XM10 camera (1376 × 1032 imaging array, 6.45 × 6.45-μm pixel size, and 14-bit digitization). Unless otherwise noted, all images and subsequent analyses were performed in the corticostriatal region (representative of the middle cerebral artery territory) using the Nikon Elements software, with images acquired using a 10X objective lens.

To analyze tomato-lectin and CD13 stains, we assessed proportional area of colocalization between TL+ staining and CD13+. We assessed the corticostriatal regions in 3 sections per slide using images from a 10X objective lens (5056 × 2960 pixels). Using Nikon Elements, the number of pixels that express both the TL+ and the CD13+ stains were averaged across all sections analyzed per animal, with average values reported within each group.

Statistical analysis

Sample sizes were formulated from power analyses based on the magnitude of differences detected in previous literature and preliminary studies. Effect sizes were >0.45 with histological changes and biochemical changes being larger than behavioral effects. Thus, we powered the study to detect differences with a β >0.8. Behavioral measures, infarct size, and immunohistochemistry were assessed by a two-way ANOVA (injury X treatment). All significant overall results (p<0.05) were followed up with a Tukey HSD post hoc analysis. Mice of both sexes were utilized in these studies, but as no sex differences were detected, all data are collapsed across sexes.

Results:

Both mTBI and imatinib reduces pericyte association with vasculature

To determine whether mTBI induces pericyte detachment, mice underwent a mild closed head TBI (or a control injury, SHAM) and were returned to their home cages for recovery for one week. Prior to surgery, mice were divided into treatment groups, with half receiving daily injections of imatinib and the other half receiving a vehicle solution daily for one week, beginning immediately after completion of the mTBI/Sham surgery. Brain tissue was collected on week later and pericyte association with the cerebrovasculature was assessed immunohistochemically with tomato lectin and CD13. As predicted, TBI (F_1,47 = 32.178, p<0.001) and imatinib (F1,47 = 10.266, p = 0.003) decreased pericyte coverage of the cerebrovasculature, as determined by colocalization of TL+ and CD13+ staining in both sexes compared to shams (Figure 2). Interestingly, while the combination of TBI and imatinib treatment decreased coverage compared to shams (p=0.004), it did not further reduce the coverage compared to injury or treatment alone. Of note, total pericyte staining was analyzed and no differences were detected across manipulations (p>0.05, data not shown).

Fig. 2.

Injury and imatinib treatment each independently led to decreases in vascular pericyte coverage in colocalization of CD13 (green) and tomato lectin (red), while combination of mTBI and treatment did not further decrease coverage. Representative images of staining; Sham injured animals (A-C), Imatinib treated (D-F), mTBI (G-I), or mTBI-Imat animals (J-L), with quantification of colocalization (M); p<0.05 vs shams, means (+/− SEM).

In order to assess whether reduced pericyte coverage recovers over time after injury, we induced an injury or sham injury and collected tissue 7-, 14-, or 28-days DPI before measuring pericyte coverage as above. mTBI decreased pericyte coverage of the vasculature at both 7 and 14 DPI, but not 28 DPI, displaying a main effect of injury (F_1,57 = 12.236, p<0.001) as well as an interaction of DPI and injury (F_5,57 = 16.172, p<0.001) (Figure 3), suggesting that mTBI significantly but transiently reduces pericyte interactions with the vasculature.

Fig. 3.

Pericyte coverage of cerebrovasculature via colocalization of CD13 (green) and tomato lectin (red) drops following mTBI out to one-month post-injury. Representative images of corticostriatal regions in Shams (A-C) and injured mice(D-F) at different time points: 7 days post-injury (DPI) (A,D), 14 DPI (B,E), and 28 DPI (C,F). Quantification of CD13+ vascular area is shown in G, p<0.05 vs shams, means (+/− SEM).

Pharmacological pericyte detachment does not worsen ischemic infarcts, but leads to motor deficits.

Next, we assessed whether pharmacological pericyte detachment would alter the effects of mTBI on ischemic stroke outcomes. We injured mice and administered imatinib post-TBI daily for one week before inducing a one-hour tMCAO. Three days after stroke induction, brains were collected via rapid decapitation for 2,3,5 triphenyltetrazolium chloride (TTC) staining to measure infarct volume. Consistent with previous experiments in our lab (Weil, 2021), we found that prior mTBI increased ischemic infarct size (F_7,59 = 9.520, p = 0.003), and there was a main effect of imatinib treatment (F_7,59 = 5.029, p = 0.029) and an interaction of injury*treatment (F_7,59 = 5.217, p = 0.027) (Figure 4A–B). These effects were driven in large part by the stark rise in infarct in the TBI-only mice.

Figure 4.

mTBI increases infarct volume following MCAO compared to sham mice, while imatinib treatment had no impact on infarct. A) Representative images of ischemic infarcts after TTC staining, with B) graphical depiction of infarct volume changes, and C) quantification of motor deficits on rotarod. p<0.05 vs shams, means (+/− SEM).

Functional outcomes were measured in these mice by utilizing the rotarod test for motor deficits. After stroke, there was an interaction of injury by treatment (F_1,57 = 10.928, p=0.002) leading to the TBI-only mice falling off the rotarod significantly faster than Sham. Additionally, the Imatinib-only mice had significantly lower latency to fall compared to Sham (p = 0.041) (Figure 4C).

Discussion

The neurovascular unit is critical for controlling blood flow and injury-induced damage (i.e., TBI or ischemia) can render the brain vulnerable to secondary degeneration. The response of pericytes depends on the type and severity of injury, with mild-moderate TBI leading to pericyte detachment away from the cerebrovasculature [27, 28]. In contrast, more severe injuries (severe TBI or ischemic stroke) can lead to pericyte constriction and vasospasm [29–31]. Here we hypothesized that mTBI would induce pericyte detachment and that mice with reduced pericyte coverage would have poorer ischemic outcomes.

There is established literature showing that TBI, even when mild in nature, negatively impacts function of the cerebrovasculature as a whole and more specifically the function of endothelium and pericytes, a cell type that has been clearly established to control contraction of small cerebral blood vessels [7, 14–17]. Given the epidemiological evidence that TBI leads to increases in long-term vascular impairments that can result in ischemic stroke or vascular contributions to cognitive impairments and dementia (VCID) [4, 5, 9, 14], it is important to determine the impact of injury on vascular damage and how TBI can be responsible for impairments in endothelial-pericyte coverage.

Clearly TBI has the potential to exacerbate subsequent ischemic outcomes via a multitude of overlapping mechanisms. Here, we attempted to replicate the pericyte-specific effects of TBI with imatinib in the absence of other, potentially confounding, effects of brain trauma. We report that imatinib reduced pericyte coverage of cerebral vessels to a level that was grossly similar to mTBI levels (Fig. 2). Interestingly, both pharmacological and mTBI-induced pericyte detachment was partial and did not result in complete loss of pericyte coverage. This is consistent with literature suggesting that mild injuries result in an acute and partial loss of pericyte coverage on the vasculature [32], indicating that pericytes may maintain some regulatory control of CBF after injury.

Pericyte detachment results in the loss of autoregulation and can lead to both arterial stiffening and loss of vasocontractile capacity [9, 20, 33] which over time can result in atherosclerotic plaque buildup and microthrombosis, both of which contribute to increased future risk of ischemic stroke [34, 35]. Pericyte loss also has been implicated in a number of pathophysiological states such as hypertension, small vessel disease, and Alzheimer’s Disease, all of which are risk factors for ischemia [36]. These consequences of pericyte detachment hint at a possible explanation for some of the mTBI-induced increases in ischemic damages.

Interestingly, there is a significant gap in the literature regarding the turnover and proliferation of pericytes in mature rodents. mTBI-induced pericyte detachment was apparent for at least two weeks post-injury, but by 28 DPI the population of pericytes covering the vasculature had returned to sham levels. While the ‘recovery’ is expected at some time, there are still questions as to the mechanism that drives these changes in cerebrovascular pericyte coverage. There are three potential explanations for the dynamic adjustments in pericyte coverage of the vessels. First, existing pericytes that detached following injury returned to vascular walls; second, reduced pericyte coverage could reflect morphological adjustments in existing cells that gradually reversed over the course of 28 days; or third, new populations of cells could populate vascular walls after previous loss of coverage. Pericytes appear to differentiate not only from mature pericytes, but also from neural crest cells, stem cells, and oligodendrocyte progenitor cells (OPCs) [29, 37]. There is also relatively little clarity in the literature as to how or when pericytes turn over in healthy or in diseased states, although it is known that pericytes can proliferate and produce immature pericytes for future development [38, 39]. Despite the return of pericyte coverage to sham levels, we cannot conclude that there is an increase in the overall number of pericytes, nor can we determine whether the pericytes that are attached to the vessels are fully functional. Future experiments would be needed to identify numerical changes in pericyte populations or functionality of these cells.

Our lab has previously shown that mTBI is responsible for cerebrovascular dysfunction as well as increased ischemic infarct size after tMCAO [1]. In this experiment, we investigated the role of pericyte detachment after mTBI on ischemic stroke, both in terms of functional outcomes and infarct volume. Notably, while we replicated our previous finding of increased ischemic infarcts in injured mice, we did not detect a similar increase in tissue damage with imatinib-induced pericyte detachment alone (Figure 4A). This indicates that pericyte detachment per se is not sufficient to mediate the exacerbation of ischemic damage following tMCAO in mice. Vascular injury is clearly a multifactorial risk factor for increased tissue damage after ischemia, and no single cell type is solely responsible for the larger infarcts after transient ischemia.

Also of note, the mice that received both mTBI and imatinib had infarct volumes that were comparable to non-injured mice or imatinib-only mice (Figure 4A). These data indicate that pericyte detachment could be protective in some capacity against the development of ischemic infarcts. The combination of injury and treatment resulting in similar levels of infarcts as the imatinib-only group signals a possibility for secondary pericyte functions to prevent larger tissue damage in the context of ischemia. Indeed, there is evidence that pericytes can adopt a phagocytotic phenotype and preventing more widespread inflammation and further vascular damage [40–42]. Thus, it seems that the detachment of pericytes could lead them to act as antigen presenting cells in an attempt to limit inflammatory damage after the initial hypoxia of an ischemic blockage.

When we measured functional outcomes after ischemia via the rotarod test, we observed impairments in motor coordination in both mTBI-only mice as well as imatinib-only mice compared to sham animals (Fig. 4C). This signals that pericyte detachment does have deleterious effects beyond only compromised vascular contractile capacity. Interestingly, similarly to the infarct volume, the combination of injury and treatment resulted in a recovery of this injury or treatment only effect, with the latency of these mice to fall off the rotarod reflecting the levels seen in sham mice. It raises questions as to why imatinib treatment would result in a clear functional deficit while the obvious gross tissue damage with TTC staining would not be apparent, and could potentially indicate a role for inflammatory or immune function in mediating tissue death while not being sufficient to prevent overall locomotor deficits.

It is likely that the mTBI-induced functional deficits and tissue damage comes in large part due to the vascular damage that limits metabolic delivery to neurons and glia that die off or are impaired after ischemia. In contrast, the imatinib-induced pericyte detachments leads to more unclear results, with motor dysfunction following tMCAO not correlating with any obvious gross damage of the brain tissue via TTC staining. It is entirely possible that imatinib, while used in other contexts as an anti-cancer therapeutic, may have other off-target effects. It has been shown that imatinib acts primarily to inactivate PDGFR-ß [21, 22], but it could be that providing this imatinib treatment may target and impair other tyrosine kinases that may result in some general locomotor dysfunction.

It is worth making note of the limitations of this study. First, imatinib was given to induce pericyte detachment in both non-injured and injured animals. As previously noted, it is possible that imatinib had off-target effects beyond the intended removal of pericytes from the vasculature. We optimized the dose and route of delivery based on existing literature and preliminary studies to prevent excessive off-target effects. It may be worth following up in future studies how pericyte detachment may also impact other glial cells (e.g., microglial or astrocytic activation) that interact with the cerebrovasculature and how that may impact the amount of post-injury tissue damage and death. A second potential limitation is based on the ischemic stroke induction timepoint of 7 DPI. While this timepoint has consistently produced increased vulnerability to stroke in TBI mice in our lab, it does not fully recapitulate more persistent brain injury-induced changes in brain and vascular functionality that led to the long-term stroke risk in the clinical population. A previous publication in our lab addresses changes in mTBI-induced stroke vulnerability over time [2].

In summary, here we establish a role of mTBI in disruption of pericyte coverage of the cerebrovasculature. Of note, this pericyte detachment persists after even a mild brain injury for at least two weeks and is not recovered to sham levels until 28 DPI. We provide evidence that this pericyte detachment is comparable to pharmacological detachment via imatinib treatment. Moreover, we have replicated previous reports that stroke-induced ischemic volume and locomotor deficits are significantly exacerbated by a pre-existing mTBI; however, pericyte detachment alone was not sufficient to increase ischemic volume. Taken together, these data suggest that while pericyte detachment is a component of vascular dysfunction after mTBI, this dysfunction on its own is not the lone determinant for the worsening of post-ischemic outcomes, necessitating a broader search for causes of mTBI-induced cerebrovascular impairments.

Highlights:

• Mild closed head injury induces pericyte detachment away from cerebrovasculature

• mTBI-induced pericyte detachment persists over time at least two weeks post-injury

• mTBI causes worsened ischemic stroke outcomes

• Pharmacological pericyte detachment does not replicate these stroke outcomes