Cortical contusion injury disrupts olfactory bulb neurogenesis in adult mice
© Radomski et al.; licensee BioMed Central Ltd. 2013
Received: 9 July 2013
Accepted: 8 November 2013
Published: 13 November 2013
Skip to main content
© Radomski et al.; licensee BioMed Central Ltd. 2013
Received: 9 July 2013
Accepted: 8 November 2013
Published: 13 November 2013
Experimental brain trauma activates quiescent neural stem cells (NSCs) to increase neuronal progenitor cell proliferation in the adult rodent brain. Previous studies have shown focal brain contusion in the form of a unilateral controlled cortical impact (CCI) stimulates NSCs to bilaterally increase neurogenesis in the adult hippocampus.
In this study we clarified the bi-lateral effects of a unilateral CCI on proliferation in the subventricular zone (SVZ) NSC niche and on neurogenesis in the olfactory bulb of adult mice. By varying the depth of impact from 1 mm to 2 mm depth, we show CCI to the left somatosensory cortex resulted in graded changes in mouse behavior and cellular pathology in the forebrain. As expected, contusion to the sensorimotor cortex resulted in motor coordination deficits in adult mice. During the first 3 days after injury, CCI increased proliferation in the impacted cortex, deeper striatum and SVZ of the forebrain ipsilateral to the CCI. In each of these regions proliferation was increased with increasing injury severity. At 30 days post-procedure, CCI resulted in a significant reduction in neurogenesis in the olfactory bulb ipsilateral to the CCI. Olfactory avoidance testing indicated disruptions in olfactory bulb neurogenesis were associated with impaired olfactory discrimination in mice post-injury.
The data demonstrate a focal cortical contusion injury to the left somatosensory cortex disrupts SVZ-olfactory bulb neurogenesis and impairs olfactory discrimination and motor coordination in adult mice.
Traumatic brain injury (TBI) initiates a complex cascade of cellular events in the adult brain resulting in cognitive and behavioral deficits . Significantly, TBI is generally followed by a spontaneous recovery process that diminishes these functional deficits over time . The partial or complete recovery of function underlies the plasticity of the brain and functional imaging studies have confirmed the ability of neural circuits to reorganize post-injury to recapitulate function . Although it remains debatable whether functional recovery involves the production of new neurons in the adult brain , it is established that many forms of experimental brain trauma activate quiescent neural stem cells (NSCs) to increase neuronal progenitor cell proliferation in the adult rodent brain [5–8].
Controlled cortical impact (CCI) is a graded, focal contusion model of experimental TBI commonly used in rodents . CCI in adult mice activates NSCs to increase proliferation in both the hippocampus and SVZ neurogenic niches [8, 10]. Cell lineage studies using transgenic mice indicate CCI evokes a sustained increase in neurogenesis in the adult hippocampus and exerts differential effects in hippocampal NSCs depending on their state of lineage progression. CCI depletes intermediate progenitor cell (IPC) populations and simultaneously activates radial glial-like (RGL) nestin-expressing type-1 multi-potent NSCs to divide and repopulate the lost IPCs . Targeted ablation of dividing nestin-expressing RGL cells inhibits the repopulation of IPCs following CCI and exacerbates the cognitive effects of CCI in mice, revealing a functional role for injury-induced neurogenesis in the hippocampus .
The effects of focal head trauma in the rodent SVZ are less defined . There is general agreement trauma to the cerebral cortex results in transient increases in cell proliferation in the adult SVZ . Whether this occurs in the SVZ of both cerebral hemispheres remains uncertain [8, 12, 13] and it is unclear what the effects of TBI-induced changes in the SVZ have on neurogenesis in the adult olfactory bulb [14–16]. Neuroblasts have been reported to migrate ectopically from the SVZ to the damaged cortex in adult TBI mice at the expense of the olfactory bulb [8, 17] but this phenomenon has only been observed transiently in the acute (3–4 days) post-injury period and was not reported to result in sustained neurogenesis deficits in the adult olfactory bulb .
In this study we sought to clarify the effects of focal cortical injury on neurogenesis in the adult olfactory bulb. We examined the bi-lateral effects of graded, unilateral CCI on proliferation in the SVZ and neurogenesis in the granular cell layer of the olfactory bulb using unbiased stereological counting techniques. Our data reveals CCI in adult mice significantly increases proliferation in the SVZ of the cerebral hemisphere ipsilateral to the CCI in the acute (3-day) post-injury period. Proliferation increased with increased CCI severity in both the ipsilateral and contralateral SVZ but was statistically significant for only the severest 2 mm depth CCI ipsilateral. In contrast to increases in SVZ proliferation, 2 mm and 1 mm depth CCI resulted in a significant decrease in neurogenesis in the granule cell layer of the olfactory bulb ipsilateral but not contralateral to the CCI one month post-injury. Olfactory avoidance testing reveals CCI results in acute deficits in olfactory behavior in mice. This behavioral deficit diminishes with time post-injury at a rate relative to CCI severity with deficits persisting in mice subjected to the severest CCI one month post-injury. These data demonstrate focal cortical contusion disrupts olfactory behavior and impairs SVZ neurogenesis to the olfactory bulb in the cerebral hemisphere ipsilateral but not contralateral to a unilateral CCI in adult mice.
Experiments were performed on 8–10 week-old adult C57BL/6 J male mice obtained from The Jackson Laboratory, Bar Harbor, ME. All procedures were approved by the Uniformed Services University of the Health Sciences Institution for Animal Care and Use Committee (IACUC). Mice were housed in the Uniformed Services University’s Center for Laboratory Animal Medicine on a standard 12 hour light–dark cycle, with feed and water available ad libitum.
Mice were placed under anesthesia in an induction chamber containing a mixture of O2 and isoflurane (2-4%) delivered by a vaporizer. Anesthesia was maintained during procedures by a mixture of O2 and 0.25-2% isoflurane using a nosecone (Dan Kopf Instruments). Mice were checked for adequate anesthesia by lack of response to toe-pinch. Core mouse body temperature was monitored by a thermometer probe and maintained at 36–38°C on a warming pad. The head was placed in a stereotactic device and held in a horizontal plane by ear and incisor bars (Stoelting). The fur was removed from the intended impact area of the cranium using an electric razor. The shaved scalp was sterilized with Betadine and 70% ethanol and the skull exposed using a small ≈ 10 mm surgical incision over the scalp. A 5.0 mm burr hole was drilled into the skull with a hand-held trephine to expose the dura mater. The impact tip (3 mm in diameter) was gently lowered to the surface dura to register 0 mm depth on the CCI electromagnetic controller (Impact One, Leica). A single contusion of 1 mm or 2 mm depth was then made to the somatosensory cortex (coordinates 0 mm Bregma, 2 mm lateral left) at an impactor velocity of 1.5 m/s with a dwell time of 100 ms and an impact angle of 15º. Sham treated animals were given a 5.0 mm burr hole craniotomy as described but not impacted. The incision was closed with sutures following CCI or sham procedure. Naïve animals were anesthetized as described for those undergoing surgery but were not placed in the stereotaxic frame and did not undergo scalp incision or craniotomy. After the procedure, the animals were returned to a cage placed on a heating pad and monitored continually until they were alert and ambulating. Recovery time was typically less than 10 minutes. Mice were housed singularly after the procedure and did not receive analgesics that could confound post-procedure behavioral measures.
Mice were given intraperitoneal injections of 100 mg/Kg body weight BromodeoxyUridine (BrdU) in saline. A single BrdU injection was administered at 10 minutes and at 24, 48 and 72 hours post-CCI/sham procedure. Naïve mice were injected at equivalent times. Mice analyzed for acute proliferative responses in the SVZ and forebrain parenchyma were perfusion-fixed 2 hours after the last BrdU injection (74 hours post-CCI/sham procedure). Mice analyzed for chronic cell fate in the olfactory bulb and forebrain were sacrificed by perfusion-fixation at 30 days post-CCI/sham procedure.
Baseline measurements were conducted for all mice used in behavioral studies 24 hours before the CCI/sham procedure followed by testing on days 1, 10 (except for rotarod test) and 29 post-CCI/sham. Naïve animals were tested on equivalent times. All behavioral testing was performed by a single investigator blinded to injury severity. Open field test. Gross motor ability and exploratory behavior was assessed using the open-field test. The mouse was placed in the center of a square 40 cm × 40 cm open field apparatus with black opaque walls (Stoelting) and permitted to explore the surroundings for 10 minutes. An overhead camera linked to a computer with ANY-maze behavioral tracking software (Stoelting) tracked and recorded the mouse’s movements and reported the total distance traveled during the test session for each mouse. Accelerating Rotarod test. The rotarod test was used to evaluate sensorimotor deficits. Mice were acclimated to the test by being placed on a rotating rod (Ugo Basile) for 240 seconds at a constant speed of 4 rpm prior to each testing session. Testing consisted of recording the animal’s mean latency to fall from the rotating rod set to accelerate from 4 to 40 rpm over 300 seconds. Each animal underwent 4 trials with a 3-min rest period in between each trial. Olfactory avoidance test. This behavioral paradigm has been described as a sensitive test to detect deficits in the ability of mice to sense an aversive scent, such as acetic acid . Briefly, the ANY-maze software was used to record the mouse’s movements within the open field apparatus in relation to filter paper squares (5 cm × 5 cm) soaked with either 5% acetic acid or distilled water (used as an odorless control). After habituating the mouse to the testing arena for 10 min, a filter paper scented with water was introduced to a corner of the open field square chamber and the duration each mouse spent investigating the filter paper was recorded during a 3 minute test period. The water-impregnated filter paper was then removed and after a 1 min interval it was replaced by a filter paper scented with acetic acid. The time each mouse spent investigating the odorant-scented filter paper was similarly recorded for 3 min. A positive investigation was defined by any time the mouse’s nares were in immediate proximity (<1 mm) of the filter paper. The time each mouse spent investigating the water-scented paper was subtracted from the value obtained with the acetic-acid odor to give the relative avoidance time.
Mice were placed under anesthesia (as described for CCI surgery) and transcardially perfused with ice-cold phosphate buffered saline (PBS) followed by 4% paraformaldehyde in PBS. The brain was dissected out and immersion fixed in fresh fixative overnight at 4°C. The brain was washed in PBS and embedded in agarose for sectioning on a vibrating VT-1000 vibratome (Leica). Brains were sectioned at 40 μm thickness in either the coronal or sagittal plane. Serial sections were stored in PBS in 96-well plates at 4°C until use. For anti-BrdU IHC, sections were incubated for one hour at room temperature (RT) in 2 N HCl and the acid neutralized by 0.1 M Borate buffer. For all IHCs, sections were blocked in 10% normal goat serum (NGS) in PBS for 1 hour at RT. Sections were incubated free-floating in primary antibodies overnight in PBS containing 1% NGS, 0.2% Triton X-100. Details of the primary antibodies used are as follows: (1) 1: 500 dilution of rat monoclonal anti-BrdU (Accurate Chemical & Scientific Corp.); (2) 1: 1,000 dilution of rabbit polyclonal anti-Glial Fibrillary Acidic Protein (anti-GFAP, Millipore); (3) 1: 1,000 dilution of rabbit polyclonal anti-Iba1 (Wako Pure Chemical Industries Ltd.); (4) 1: 500 dilution of mouse monoclonal anti-NeuN (clone A60, Millipore). Primary antibody combinations for dual IHC were incubated and processed simultaneously. Immunolabeled sections were washed in PBS and incubated with Alexa Fluor ® 488, 555 or 647 dye-labeled secondary antibodies (Life Technologies) at 1:1,000 dilutions in PBS for 60 minutes at RT, counterstained with DAPI and mounted under Mowiol. Immunofluorescent images were captured on a Zeiss A1 or M1 (stereology only) Imager microscopes with Axiocam digital cameras, or on a Zeiss Pascal laser scanning confocal microscope.
Unbiased stereological methods were conducted on blind-coded slides by a single investigator blinded to the injury status of the animals. All stereological cell counts were performed using MicroBrightField Bioscience Stereo Investigator software linked to a Zeiss M1 imager microscope with a motorized x, y and z-plane stage. The total number of anti-BrdU immunolabeled cells (BrdU+) cells in the SVZ and dual anti-BrdU/anti-NeuN (BrdU+/NeuN+) immunolabeled cells in the granule cell layer of the olfactory bulb were estimated using the unbiased Optical Fractionator probe of the StereoInvestigator program. Stereological cell counts were obtained from both cerebral hemispheres ipsilateral and contralateral to the CCI/sham (a single hemisphere was counted for naïve controls). To determine the number of BrdU immunolabeled cells in the lateral wall of the SVZ, 6 sections (240 μm apart), starting at approximately 1.46 mm rostral to Bregma, were selected from each animal and processed for anti-BrdU IHC as described. Systematic random sampling areas (137 μm × 104 μm) in each section through the anterior portion of the SVZ were randomly chosen by the StereoInvestigator program with a counting frame area of 50 μm × 50 μm. Section thickness was determined by marking when DAPI-counterstained cells first appeared in focus. Z series of 30 μm depth with 1 μm intervals between images were collected. BrdU+/DAPI + nuclei that first appeared in focus within the 20 μm Z-stack counting probe were counted. In the olfactory bulb, the number of BrdU/NeuN double-positive cells in the granular cell layer was counted on 5 sagittal sections (160 μm apart). Positive profiles were counted within a 60 μm × 60 μm counting frame spaced in a 300 μm × 300 μm sampling grid with a 30 μm Z-stack counting probe. Dual anti-BrdU/anti-NeuN immunolabeled cells were defined by co-localization of the two immunofluorescent signals in the same z-plane. All together, the Gundersen (m = 1) coefficient of error ranged between 0.06 and 0.12.
Counts of dual anti-BrdU/anti-GFAP, dual anti-BrdU/anti-Iba1, or dual anti-BrdU/anti-NeuN cells were made from digitally captured images from peri-contusional areas of the neocortex and the striatum. Counts were made from the cerebral hemisphere ipsilateral to the CCI or sham only as minimal anti-BrdU labeling was observed in non-neurogenic regions in the contralateral cerebral hemisphere (as well as from a single cerebral hemisphere in naïve mice). In each case, z-stacks or wide-field z-plane series fluorescent images were collected at 20× magnification from peri-contusional areas and the number of dual immunofluorescent cells present within the image area recorded. Cells were counted if the whole nucleus was inside the image area.
For analysis of lesion volume, nine coronal sections at 440 μm intervals spanning the extent of the injury were processed for DAPI immunohistochemistry and imaged using a Hamamatsu Nanozoomer 2.0-RS digital slide scanner. Measurements of cortical lesion area in each section were performed using the ImageJ software. The total volume of the cortical lesion was calculated by multiplying the area obtained from each section by the distance between each section.
All histograms are plotted as mean plus standard error of the mean (SEM) error bar above. Multiple comparisons were analyzed using repeated measures one-way ANOVA with post hoc Tukey’s test unless otherwise stated. P-values < 0.05 were considered significant. Unless otherwise stated, p-values stated are from Tukey’s test comparisons to naïve controls.
To confirm these results were specific to a lack of coordination and not a paucity of movement in mice, we used the open field test to measure unconditioned motor activity. This test revealed CCI did not result in acute differences in locomotor activity when compared to naïve mice at 1 day post-procedure (Figure 1b). However, when compared to sham controls at this acute post-injury stage, moderate CCI mice showed increased locomotion (total distance traveled) (p < 0.05, Figure 1b). This trend of increased activity persisted and total locomotor activity significantly increased with mild and moderate CCI at 10 days (p < 0.01 and p < 0.001 respectively, Figure 1b) and 29 days post-procedure (p < 0.001 and p < 0.05 respectively, Figure 1b). These results demonstrate CCI to the somatosensory cortex results in acute deficits in motor coordination proportional to contusion depth and sustained locomotor hyperactivity in mice. Combined these tests confirm our CCI paradigms result in graded acute and chronic behavioral changes in adult mice.
Quantification of the IHC data confirms the majority of post-injury proliferating cells in the peri-contusional areas of the ipsilateral cortex and striatum were Iba1-expressing microglia/macrophages, with minimal contribution to the proliferating cell pool from GFAP-expressing astrocytes (Figure 2f). Of note, quantification revealed a graded response to injury in the ipsilateral cerebral hemisphere with the density of proliferating BrdU + cells increasing with increasing injury severity from sham to mild to moderate CCI (Figure 2f). Hence, both behavioral and histological assessments of our CCI/sham mice reveal graded functional and histopathological changes in response to increasing injury severity.
Neuroblasts generated in the SVZ migrate to the olfactory bulb along the rostral migratory stream (RMS) where they differentiate into interneuron subtypes . In adult mice, the vast majority of surviving SVZ neuroblasts differentiate into granule cell interneurons in the deeper granular cell layer [20, 21]. Accordingly, we used stereological cell counting methods to quantify the numbers of BrdU-pulse labeled neurons in the granular cell layer identified by dual anti-BrdU/anti-NeuN IHC. Mice were injected with BrdU at 0, 24, 48 and 72 hours post-procedure as before and processed for dual IHC 30 days post-procedure.
Cell counts in the ipsilateral cortex reveal reductions in olfactory bulb neurogenesis were associated with a corresponding increase in the presence of BrdU + cells in peri-contusional cortical areas in CCI mice at 30 days post-procedure (p < 0.05, Figure 5d). Co-immunolabeling with anti-NeuN antisera revealed negligible numbers of dual NeuN+/BrdU + cells (< 1% of BrdU+) in the cortex of all animal groups (Figure 5e). In contrast, BrdU + cells in the ipsilateral cortex of CCI mice coimmunolabel with anti-GFAP or anti-Iba1 antisera (Figure 5f), confirming a non-neuronal fate of many BrdU + cells at 30 days post-injury. Cell counts reveal more BrdU + cells are GFAP + (35.3 ± 9.9 in mild CCI and 43.3 ± 4.6 in moderate CCI mice) than Iba1+ (6.2 ± 5.1 cells in mild CCI and 17.7 ± 4.3 in moderate CCI mice) in CCI mice at this stage (Figure 5g). These results indicate a significant fraction of the surviving cells labeled with BrdU during the acute post-injury period have acquired an astrocytic fate by 30 days post-injury.
Taken together, our analyses demonstrate unilateral, focal contusion to the somatosensory cortex impairs sensorimotor coordination and olfactory avoidance behavior in adult mice. These behavioral changes are accompanied by increased cell proliferation and decreased olfactory bulb neurogenesis in the ipsilateral cerebral hemisphere.
In this study we demonstrated that a unilateral contusion to the left somatosensory cortex decreased neurogenesis in the olfactory bulb of the cerebral hemisphere ipsilateral to the injury. Decreases in olfactory bulb neurogenesis were recorded in mice subjected to a moderate 2 mm and mild 1 mm depth controlled cortical impact (CCI) at 30 days post-CCI. Chronic deficits in olfactory bulb neurogenesis induced by CCI were in contrast to the acute pro-proliferative effects of CCI in the SVZ neurogenic niche at 3 days post-CCI. Hence combined stereological analysis of adult SVZ and olfactory bulb cell populations reveal CCI is capable of stimulating proliferation in the SVZ neurogenic niche but this increase does not result in a corresponding increase in olfactory bulb neurogenesis.
One explanation for the failure of injury-induced neuroblasts to increase olfactory bulb neurogenesis is that cortical contusion disrupts the normal maturational process from the SVZ to the olfactory bulb. Neuroblasts born in the adult SVZ undergo chain migration within the RMS en route to the olfactory bulb where they switch to radial migration to reach destinations in the granular cell or periglomerular layers to form GABAergic interneurons [22, 23]. Gross interruptions in the RMS by the CCI lesion cavity could account for the failure of migrating neuroblasts to reach olfactory bulb targets in moderate CCI mice in which the contusion resulted in a cavity extending through the RMS into the lateral ventricle in all mice analyzed at 30 days post-injury. However, the lesion was not as consistently extensive in mild CCI mice at 30 days post-injury – in almost half of the mice analyzed the lesion was confined to the cortex, sparing the corpus callosum and underlying RMS - and yet mild CCI mice also exhibited substantial deficits in neurogenesis at this stage (1.63-fold reduction compared to a 1.87-fold reduction with moderate CCI). This suggests additional factors such as cell death from the toxic effects of neuroinflammation  or the ectopic recruitment of SVZ neuroblasts [25–28] to damaged cortical tissue areas contributed to the reduction in olfactory bulb neurogenesis.
Similarly, behavioral testing demonstrated CCI resulted in a corresponding impairment of olfactory avoidance behavior in adult mice. This suggests an association between CCI deficits in olfactory bulb neurogenesis and impaired olfactory avoidance behaviors in adult mice. However, impaired olfactory discrimination persisted in moderate but not mild CCI mice at 30 days post-CCI and yet neurogenesis was significantly reduced by both CCI procedures. Therefore, it is unlikely that decreases in neurogenesis can solely account for this impairment. Damage to additional structures - the olfactory nerve or olfactory processing areas such as the amygdala, hippocampus and hypothalamus - may also contribute to this behavioral deficit.
In addition to injury effects on olfactory bulb neurogenesis and olfactory-related behaviors, the graded cortical contusion injury produced here resulted in unconditioned locomotor activity, deficits in motor coordination, increased macrophage/microglia cell proliferation in peri-contusional areas and chronic astrogliosis that were commensurate with lesion severity. These results are consistent with the electromagnetic CCI device producing a reproducible and graded contusion depth and graduated outcomes. Although there is good evidence that craniotomy alone in sham mice produces injury-associated effects , we were able to distinguish between the histological effects of two contusion depths differing by 1 mm with as few as 4 mice per group using unbiased stereological cell counting procedures. In behavior studies group sizes of 10 or 12 were sufficient to identify behavioral differences between the two CCI depths. This degree of injury reproducibility is comparable to published data using the same CCI device with accelerating rotarod testing in which performance differences were distinguishable between 0.5 mm changes in CCI depth with a group size of 12 (CCI at 2.7 mm left lateral and +3 mm Lambda ≈ -1.2 mm Bregma, see ).
Recently, two studies of graded CCI severity to the left parietal lobe have revealed increasing cognitive and emotional deficits with increasing CCI severity [30, 31]. Spatial learning and memory deficits were most closely related to CCI severity in both studies, as would be expected from the location of their CCI directly superior to the hippocampus (CCI between Bregma and Lambda anatomical landmarks). In contrast, neither study observed the CCI-induced hyperactivity in open-field testing we detected. We speculate differences in locomotor behavior could result from differences in the rostral-caudal location of the CCI (left lateral at Bregma in our study versus between Bregma and Lambda). Thus, CCI studies consistently demonstrate the high comparative value of this approach in TBI research but also highlight the need for the detailed spatial definition of the contusion injury in comparative analysis.
We are grateful for assistance with surgical and behavioral procedures from Laura Tucker, Oz Malkesman and Amanda Fu in the Animal Models Core at the Center for Neuroscience and Regenerative Medicine (CNRM) at USU. We also thank Dennis McDaniel in the Bio-instrumentation Center (BIC) at USU for assistance with microscopy and stereology. This work was supported by a grant from the Center for Neuroscience and Regenerative Medicine (CNRM).
This article is published under license to BioMed Central Ltd. This is an open access article distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/2.0), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.