This study demonstrates delivery of a repetitive traumatic brain injury to mice and simultaneous implantation of a cranial window for subsequent intravital imaging of a neuron-expressed EGFP using two-photon microscopy.
The goal of this protocol is to demonstrate how to longitudinally visualize the expression and localization of a protein of interest within specific cell types of an animal’s brain, upon exposure to exogenous stimuli. Here, the administration of a closed-skull traumatic brain injury (TBI) and simultaneous implantation of a cranial window for subsequent longitudinal intravital imaging in mice is shown. Mice are intracranially injected with an adeno-associated virus (AAV) expressing enhanced green fluorescent protein (EGFP) under a neuronal specific promoter. After 2 to 4 weeks, the mice are subjected to a repetitive TBI using a weight drop device over the AAV injection location. Within the same surgical session, the mice are implanted with a metal headpost and then a glass cranial window over the TBI impacting site. The expression and cellular localization of EGFP is examined using a two-photon microscope in the same brain region exposed to trauma over the course of months.
Traumatic brain injury (TBI), which can result from sports injuries, vehicle collisions, and military combat, is a worldwide health concern. TBI can lead to physiological, cognitive, and behavioral deficits, and lifelong disability or mortality1,2. TBI severity can be classified as mild, moderate, and severe, the vast majority being mild TBI (75%-90%)3. It is increasingly recognized that TBI, particularly repetitive occurrences of TBI, can promote neuronal degeneration and serve as risk factors for several neurodegenerative diseases, including Alzheimer's disease (AD), amyotrophic lateral sclerosis (ALS), frontotemporal dementia (FTD), and chronic traumatic encephalopathy (CTE)4,5,6. However, the molecular mechanisms underlying TBI-induced neurodegeneration remain unclear, and thus represent an active area of study. To gain insight into how neurons respond to and recover from TBI, a method for monitoring fluorescently tagged proteins of interest, specifically within neurons, by longitudinal intravital imaging in mice after TBI is described herein.
To this end, this study shows how to combine a surgical procedure for the administration of closed-skull TBI that is similar to what that has been reported previously7,8, together with a surgical procedure for implantation of a cranial window for downstream intravital imaging, as described by Goldey et al9. Notably, it is not feasible to implant a cranial window first and subsequently perform a TBI in the same region, as the impact of the weight drop that induces the TBI is likely to damage the window and cause irreparable harm to the mouse. Therefore, this protocol was designed to administer the TBI and then implant the cranial window directly over the impact site, all within the same surgical session. An advantage of combining both the TBI and cranial window implantation in a single surgical session is a reduction in the number of times a mouse is subjected to surgery. Further, it allows one to monitor the immediate response (i.e., on the timescale of hours) to TBI, as opposed to implanting the window at a later surgical session (i.e., initial imaging starting on a timescale of days post-TBI). The cranial window and intravital imaging platform also offer advantages over monitoring neuronal proteins by conventional methods such as immunostaining of fixed tissues. For example, fewer mice are required for intravital imaging, as the same mouse can be studied at multiple time points, as opposed to separate cohorts of mice needed for discrete time points. Further, the same neurons can be monitored over time, allowing one to track specific biological or pathological events within the same cell.
As a proof of concept, the neuron-specific expression of enhanced green fluorescent protein (EGFP) under the synapsin promoter is demonstrated here10. This approach can be extended to 1) different brain cell-types by utilizing other cell-type specific promoters, such as myelin basic protein (MBP) promoter for oligodendrocytes and glial fibrillary acidic protein (GFAP) promoter for astrocytes11 , 2) different target proteins of interest by fusing their genes with the EGFP gene, and 3) co-expressing multiple proteins fused to different fluorophores. Here, EGFP is packaged and expressed via adeno-associated virus (AAV) delivery through an intracranial injection. A closed-skull TBI is administered using a weight-drop device, followed by implantation of a cranial window. Visualization of neuronal EGFP is achieved through the cranial window, using two-photon microscopy to detect EGFP fluorescence in vivo. With the two-photon laser, it is possible to penetrate deeper into the cortical tissue with minimal photodamage, allowing for repeated longitudinal imaging of the same cortical regions within an individual mouse for days and up to months12,13,14,15. In sum, this approach of combining a TBI surgery with intravital imaging aims to advance the understanding of the molecular events that contribute to TBI-induced disease pathology16,17.
All the animal related protocols were conducted in accordance with the Guide for the Care and Use of Laboratory Animals published by the National Research Council (US) Committee. The protocols were approved by the Institutional Animal Care and Use Committee of University of Massachusetts Chan Medical School (UMMS) (Permit Number 202100057). In brief, as shown in the schematic of study (Figure 1), the animal receives a virus injection, a TBI, a window implantation, and then intravital imaging in a time sequence.
NOTE: Commercial terms have been removed. Please refer to the Table of Materials for the specific equipment used.
1. Intracranial injection of AAV using a stereotaxic device
2. Administration of a repetitive TBI induction
NOTE: The TBI parameters are adjusted from previous reports7,8, in which the TBI impact was delivered once. The protocol here applies the same parameter, except increasing the total impact number to 10.
3. Cranial window implantation surgery
NOTE: The cranial window implantation steps below were adopted from Goldey et al.9, and their specifications of the headpost and the imaging well were applied here.
4. Intravital two-photon imaging
As proof of concept for this protocol, viral particles expressing AAV-Syn1-EGFP were injected into the brain cortex of male TDP-43Q331K/Q331K mice (C57BL/6J background)19 at the age of 3 months. It is noted that wild-type C57BL/6J animals can also be used, however this study was carried out in TDP-43Q331K/Q331K mice because the laboratory is focused on neurodegenerative disease research. A TBI surgery was performed 4 weeks after AAV injection. Within the same surgical setting, the headpost and cranial window were implanted. The mouse was serially imaged using a two-photon microscope at day 0, 1 week, and 4 months post-TBI surgery (Figure 1). In general, the injection surgery took ~30 min, and the TBI surgery took ~1 h per mouse. During the TBI surgery, the mice generally required a longer amount of time to right themselves from a supine position to a prone position after subsequent impacts, compared to the initial impacts. For example, a mouse might have required 2 min to right its position after the first impact, but required 10 min to right its position after the 10th impact. For cranial window implantation surgery, ~3 h was usually required to perform all the steps, but took longer than necessary if persistent bleeding occurred. Two-photon imaging usually required 20 min per mouse to acquire all the images. For the current study involving three mice with the expression of EGFP, the animals did not show evidence of a skull fracture, nor did they die during surgery or up until 4 months after surgery. Both the morbidity and skull fracture rate were 0%. This is consistent with the relatively low (<10%) mortality rate in a similar model of inducing TBI but without a cranial window implant7,8.
A weight-drop device (Figure 2A), that has been previously reported7,8, was used to deliver TBI impacts onto the mouse head on the right-side, as indicated in Figure 2B. A transparent plastic tube (no. 5 in Figure 2A; 60 cm in length and 1.4 cm in inner diameter) was securely and vertically clamped to a metal bracket, and a plastic impactor column (no. 7 in Figure 2A; 10 cm in length and 1.3 cm in diameter, with a flat round tip of 2 mm in diameter) was placed in the vertical tube. A 50 g metal weight (no. 6 in Figure 2A) was placed above the impactor and tethered with a nylon string (no. 4 in Figure 2A). A buffer cushion (no. 8 in Figure 2A; 9 cm x 9 cm x 1 cm [length x width x depth], made of polyethylene foam and cotton gauze) was placed under the mouse head to buffer and absorb the impact energy.
The cranial window was made of two glass coverslips combined by optical adhesive (Figure 2C), and the glass coverslip of 3 mm diameter was the side touching the brain surface. Two-photon imaging was carried out at two different depths (Figure 3A) from the brain surface: 1) a superficial level where the cortical vasculature was abundant and had a relatively large lumen diameter that appeared black, but with sparse EGFP positive cell bodies; and 2) a deeper level where the vasculature was sparse and had a small lumen diameter, with many EGFP positive cells visible.
At ~4 h (day 0) after surgery, two-photon imaging was carried out at three planes, with an interval of 10 µm at the superficial level (layer I, less than 100 µm from the meningeal surface) first where the vasculature appeared black (indicated by the dashed red lines in Figure 3B), and was used as a reference map for the next imaging session to locate the original imaging region. At this superficial level, the cortical vasculature and neuronal processes were the predominate structures that could be observed, whereas the EGFP positive cell bodies were sparse. After imaging within the superficial level, the focus was adjusted downward by ~400 µm, deeper than the superficial level, into the cortex (layer IV and V) to image the EGFP positive neuronal cell bodies. Images were acquired within six planes, with an interval of 10 µm. EGFP protein expression was diffusely distributed throughout the cell body, as indicated in Figure 3C. There was occasional occurrence of some EGFP puncta outside the cell bodies, which could represent EGFP inclusions that formed over time within axons or dendrites. This protocol results in AAV-SYN1-EGFP expression ~2-3 mm surrounding the injection site, which can be defined by histological analysis of post-mortem tissues.
At 1 week and 4 months post-TBI, the vasculature pattern observed at the day 0 time point was used as a reference to locate the same imaging region (Figure 3D,F). The vasculature pattern in Figure 3D,F is similar to that in Figure 3B. The imaging procedure for the 1 week and 4 months post-TBI was the same as that described above for the day 0 time point, and EGFP expression was detected throughout the cell body as before (Figure 3E,G). The fluorescence intensity at day 0 and 4 months were similar at both the superficial and deeper levels. However, the fluorescence intensity at 1 week was lower than that of day 0 and 4 months at both the superficial and deep levels, which could be due to the translational repression reported for other TBI models20. Notably, the quality of the images at 4 months compared to the day 0 time point demonstrates that the cranial window has maintained clarity and integrity, and that the viral expression is still robust 4 months post-surgery for effective intravital imaging. As EGFP expresses as a relatively diffuse protein, the fluorescence signals may be better resolved and discreet when EGFP is fused to a protein of interest.
Figure 1: The timeline for this intravital imaging protocol. The protocol is initiated with virus injection. TBI and cranial window surgery are performed within the same surgery session at 2-4 weeks after virus injection. Intravital imaging is carried out at day 0, 1 week, 4 months after TBI. Please click here to view a larger version of this figure.
Figure 2: Diagrams for the TBI device, TBI impact site, and window preparation. (A) Equipment for the weight-drop TBI device. 1: pedestal, 2: bracket, 3: adjustable clamp, 4: nylon string, 5: weight dropping tunnel, 6: metal weight column (50 g), 7: impactor, 8: buffer cushion. (B) A schematic of the virus injection and TBI impact site, with the coordinates 2.5 mm posterior to the Bregma and 2 mm lateral from the sagittal suture to the right. (C) A schematic for preparing the glass window. Two glass cover slips, 3 mm and 5 mm in diameter, are combined using optical adhesive. (D) Pictures of the metal headpost and imaging well. Please click here to view a larger version of this figure.
Figure 3: Schematic of imaging planes and the image data. (A) Multiple planes are imaged at a superficial level where the vasculature is located and a deep level. The images are collected as follows: three planes at the superficial level, where the cortical vasculature and neuronal processes are the predominate EGFP-positive structures, and six planes at the deep level, where EGFP-positive cell bodies are predominately observed, with a 10 µm interval between the neighboring planes. Each plane is of a 425.10 µm x 425.10 µm size. (B) Representative image of a plane in the superficial level at the day 0 time point. The dashed red lines indicate the vasculature outline that can be used as a reference to locate the original imaging place for subsequent time points. (C) Representative image of a plane in the deep level at the day 0 time point soon after TBI. (D) Representative image of a plane in the superficial level at the 1 week time point. The dashed red lines denote the vasculature, similar to the outline on day 0 in Figure 3B, confirming that the two-photon imaging was carried out at a similar location on day 0 and 1 week after TBI. (E) Representative image in the deep level at the 1 week time point post-TBI. (F) Same as B and D, at 4 months post-TBI. (G) Same as C and E, at 4 months post-TBI. Scale bar: 50 µm. Please click here to view a larger version of this figure.
In this study, AAV injection, TBI administration, and a headpost with cranial window implantation were combined for longitudinal imaging analysis of EGFP-labeled neurons within the mouse brain cortex (layers IV and V) to observe the effects of TBI on cortical neurons. This study notes that the TBI site chosen here, above the hippocampus, provides a relatively flat and broad surface for implantation of the cranial window. Conversely, the skull is relatively narrow anterior to this site, and therefore it is difficult to ensure that the headpost will effectively contact the surface of the skull. While only the TDP-43Q331K/Q331K mouse model was used in this study, considering the laboratory's research focusing on ALS and FTD19, this protocol should be applicable to most other mouse strains. In addition to labeling neurons as described here, various strategies can be used to label other cell types for intravital imaging. One approach is to express the genetically encoded fluorescent proteins in a cell-specific manner, using the Cre-Lox recombination system in genetically modified mice21. Another approach is to employ certain viral serotypes for cell-specific transduction of genetically encoded fluorescent proteins. Site-specific delivery can be achieved by performing the intracranial injection at the desired location in the brain. The efficiency and ease with which one can express cell-specific proteins via viral transduction for intravital imaging is an advantage over creating transgenic mouse models.
For the surgery protocols, there are multiple critical steps that need to be carefully performed. When drilling holes on the skull for virus injection at the start of the protocol, one needs to be cautious of damaging the tissue underneath the skull. If tissue damage occurs while drilling on the skull, there will be inflammation, tissue adhesion, and angiogenesis around the injury site, thereby increasing the risk of bleeding during the craniotomy for cranial window implantation. For administration of the TBI with the weight-drop device, the present study placed a cushion under the mouse head to buffer the TBI impact force, and thereby decrease the chance of a skull fracture. No obvious head motion on rotational directions and acceleration were observed, supporting the notion that this TBI model has a relatively low chance of diffusion axonal injury. Interestingly, Foda et al. used a similar weight-drop device to construct a closed-skull TBI injury on rats, and a larger, soft foam cushion (12 cm thick) was placed under the rat head so that it could easily move in response to the TBI impact force on rotational direction with acceleration, thereby resulting in diffusion axonal injury22. One of the advantages of the tube-guiding weight drop TBI model is that the severity of the trauma can be modulated by changing the weight drop height (i.e., a higher height for a larger force) and/or repeating the number of times the impact is delivered. For example, Flierl et al. used a similar weight-drop device as the present study to induce TBI on mice; the weight of the metal was 333 g, which induced a mild TBI when dropped from a height of 2 cm and a severe TBI when dropped from a height of 3 cm23. The craniotomy step prior to window implantation also needs to be performed carefully, to avoid damaging the brain tissue under the skull. Periodically moving the drill bit to a different region on the skull helps to avoid over-drilling at the same spot, which can result in excessive heating, tissue damage, and bleeding; further, frequently irrigating with saline can also cool the drilling site. When implanting the glass window, it is important to align the glass plane such that it is parallel to the skull surface plane; a snug window fit within the skull prevents leakage of the dental cement liquid into the space between the window and the brain.
If the window is blurry at day 0 post-TBI surgery, but fluorescent signals are detected, the dental cement may have entered the space between the window and the brain, thereby forming a cement layer that covers the brain surface and obscures the fluorescence emission. In this situation, one can remove the window and the dental cement using the drilling method described in protocol step 3.5, and then implant a new glass window at the original site, as described in detail by Goldey et al.9. If the window becomes blurry at a subsequent stage during the longitudinal imaging process, one can attempt to remove potential debris by gently rubbing the window with a cotton-tipped applicator. If this does not resolve the issue, it may be due to regrowth of the bone and the meninges underneath the glass window. In this case, one can remove the window and drill to remove the regrown bone and/or tweeze apart and remove the meninges, followed by implantation of a new glass window at the original site, as described by Goldey et al.9. As shown in the results, EGFP expression and fluorescence is robust over a time course of ~4 months post-TBI. One can expect a decrease of fluorescence signal within the first week post-TBI, which is likely due to TBI-induced translational repression that causes a temporary reduction in global protein synthesis20.
To maintain high imaging quality and achieve imaging in multiple planes, mice were under isoflurane anesthesia to limit movement. However, it is important to note that anesthesia may affect the study results. For example, it has been reported that mice under isoflurane anesthesia exhibit different microglial activity in response to photodamage compared to awake mice24. For two-photon imaging, the scanning speed and the number of scans for signal averaging (set as "number", under "averaging") are the main factors that can determine image resolution. To optimize resolution, one can reduce the scanning speed and increase the averaging number, however, a slower speed and higher averaging number increases the imaging time for a particular field of view and may cause photo bleaching. The present study aims to accomplish chronic long-term imaging on the same animal at the same brain location. To avoid potential photo bleaching while achieving sufficient resolution, two-photon imaging was carried out at a scanning speed of 8 with an averaging number of 16.
An alternative approach to performing a craniotomy and implanting a glass window is to thin the skull, to the extent that it is transparent and "paper thin" for live imaging25,26,27. The thin-skull window can maintain the intactness of the skull, and thus avoid the inflammation induced by the surgical operation. Therefore, the thin-skull window approach may be preferred for live imaging with inflammation-relevant markers and/or over a relatively shorter period of time (i.e., ~14 days post-surgery, as reported25). For long-term imaging (i.e., 4 months, as described here) of chronic neurodegenerative processes, in addition to assessing the acute effects of TBI on the same animal, it is recommended to use a glass window implanted by a craniotomy method.
As mentioned in the introduction, there are several notable advantages of intravital imaging for studying various biological and pathologic events in the mammalian brain. However, there are some limitations of this protocol as well. For example, contrecoup brain injury describes the brain contusion or hematoma remote from, usually opposite to, the force contacting site28,29,30. In the present study, the TBI impacts were delivered to the parietal lobe; contrecoup injury would be expected ventrally and inaccessible to intravital imaging. A histological analysis could be used as a complementary approach to further examine phenotypes of interest outside the impact site covered by the cranial window. In addition, exogenous overexpression of certain proteins may also induce toxicity or pathology. In the present study, some occasional EGFP puncta were observed, which may represent accumulations due to protein overexpression. Therefore, it is recommended to include a negative-control protein (e.g., EGFP or RFP alone) in the study design to address this possibility, particularly if the protein of interest is prone to forming punctate structures. The number of proteins that one can study simultaneously is limited by the lasers and capabilities of the two-photon system. It may be possible to image more than one fluorophore (e.g., EGFP, RFP, etc.) by two-photon microscopy, although multiplexing capabilities with conventional wide-field and confocal microscopes often allow analyses of more proteins within a single tissue sample. Finally, it is important to include a sham control that receives all the same procedures as the TBI animal, except the weight-drop impacts. Cranial window implantation surgery is an invasive operation and can cause some local changes, such as inflammation and altered intracranial pressure, which could affect the TBI process. A sham control helps the experimentalist assess phenotypes that are due to the impact versus the surgical procedures.
In summary, this protocol introduces a method to longitudinally image proteins specifically in the neurons of the mouse brain cortex in response to TBI. This protocol can be modified to analyze the expression and localization of various cell-specific proteins in response to TBI. The goal of this protocol is to provide an approach for studying the immediate and long-term consequences of TBI in the mammalian brain.
The authors have nothing to disclose.
We thank Dr. Miguel Sena-Esteves at the University of Massachusetts Chan Medical School for gifting the AAV(PHP.eB)-Syn1-EGFP virus, and Debra Cameron at the University of Massachusetts Chan Medical School for drawing the mice skull sketch. We also thank current and past members of the Bosco, Schafer and Henninger labs for their suggestions and support. This work was funded by the Department of Defense (W81XWH202071/PRARP) to DAB, DS, and NH.
Adjustable Precision Applicator Brushes | Parkell | S379 | |
BD insulin syringe | BD | NDC/HRI#08290-3284-38 | 5/16" x 31G |
Betadine | Purdue | NDC67618-151-17 | including 7.5% povidone iodine |
Buprenorphine | PAR Pharmaceutical | NDC 42023-179-05 | |
Cefazolin | HIKMA Pharmaceutical | NDC 0143-9924-90 | |
Ceramic Mixing Dish | Parkell | SKU: S387 | For dental cement preparation |
Cotton Tipped Applicators | ZORO | catlog #: G9531702 | |
Catalyst | Parkell | S371 | full name: "C" Universal TBB Catalyst |
Dental cement powder | Parkell | S396 | Radiopaque L-Powder for C&B Metabond |
Dental drill | Foredom | H.MH-130 | |
Dental drill controller | Foredom | HP4-310 | |
Dexamethasone | Phoenix | NDC 57319-519-05 | |
EF4 carbide bit | Microcopy | Lot# C150113 | Head Dia/Lgth/mm 1.0/4.2 |
Ethonal | Fisher Scientific | 04355223EA | 75% |
FG1/4 carbide bit | Microcopy | Lot# C150413 | Head Dia/Lgth/mm 0.5/0.4 |
FG4 carbide bit | Microcopy | Lot# C150309 | Head Dia/Lgth/mm 1.4/1.1 |
Headpost | N/A | N/A | Custom-manufactured |
Heating apparatus | CWE | TC-1000 Mouse | equiped with the stereotaxic instrument and be used while operating surgery |
Heating blanket | CVS pharmacy | E12107 | extra heating device and be used after surgery |
Isoflurane | Pivetal | NDC 46066-755-03 | |
Isoflurane induction chamber | Vetequip | 89012-688 | induction chamber for short |
Isoflurane volatilizing machine | Vetequip | 911103 | |
Isoflurane volatilizing machine holder | Vetequip | 901801 | |
Leica surgical microscope | Leica | LEICA 10450243 | |
Lubricant ophthalmic ointment | Picetal | NDC 46066-753-55 | |
Marker pen | Delasco | SMP-BK | |
Meloxicam | Norbrook | NDC 55529-040-10 | |
Microinjection pump and its controller | World Precision Instruments | micro4 and UMP3 | |
Microliter syringe | Hamilton | Hamilton 80014 | 1701 RN, 10 μL gauge for syringe and 32 gauge for needle, 2 in, point style 3 |
Mosquito forceps | CAROLINA | Item #:625314 | Stainless Steel, Curved, 5 in |
Depilatory agent | McKesson Corporation | N/A | Nair Hair Aloe & Lanolin Hair Removal Lotion |
Microscope 1 | Nikon | SMZ745 | Nikon microscope for cranial window preparation |
Microscope 2 | Zeiss | LSM 7 MP | two-photon microscope |
Multiphoton laser | Coherent | Chameleon Ultra II, Model: MRU X1, VERDI 18W | laser for two-photon microscopy |
Non-absorbable surgical suture | Harvard Apparatus | catlog# 59-6860 | 6-0, with round needle |
Norland Optical Adhesive 81 | Norland Products | NOA 81 | |
No-Snag Needle Holder | CAROLINA | Item #: 567912 | |
Quick base liquid | Parkell | S398 | "B" Quick Base For C&B Metabond |
Regular scissor 1 | Eurostat | eurostat es5-300 | |
Regular scissor 2 | World Precision Instruments | No. 501759-G | |
Round cover glass 1 | Warner instruments | CS-5R Cat# 64-0700 | for 5 mm of diameter |
Round cover glass 2 | Warner instruments | CS-3R Cat# 64-0720 | for 3 mm of diameter |
Rubber rings | Orings-Online | Item # OO-014-70-50 | O-Rings |
Saline | Bioworld | L19102411PR | |
Spring scissor 1 | World Precision Instruments | No. 91500-09 | tip straight |
Spring scissor 2 | World Precision Instruments | No. 91501-09 | tip curved |
Stereotaxic platform | KOPF | Model 900LS | |
Super glue | Henkel | Item #: 1647358 | |
surgical Caliper | World Precision Instruments | No. 501200 | |
Surgical forceps 1 | ELECTRON MICROSCOPY SCIENCES | Catlog# 0508-5/45-PO | style 5/45, curved |
Surgical forceps 2 | ELECTRON MICROSCOPY SCIENCES | catlog# 0103-5-PO | style 5, straight |
Surgical forceps 3 | ELECTRON MICROSCOPY SCIENCES | catlog# 72912 | |
Surgical forceps 4 | ELECTRON MICROSCOPY SCIENCES | Catlog# 0508-5/45-PO | style 5/45, curved |
Surgical gauze | ZORO | catlog #: G0593801 | |
Surgical lamp | Leica | Leica KL300 LED | |
UV box | Spectrolinker | XL-1000 | also called UV crosslinker |
Vaporguard | Vetequip | 931401 | |
Vetbond Tissue Adhesive | 3M Animal Care | Part Number:014006 |