Two-photon intravital imaging can be used to investigate interactions among different cell types in the spinal cord in their native tissue environment in a bone marrow chimeric animal with a dorsal column traumatic spinal cord crush injury.
Traumatic spinal cord injury causes an inflammatory reaction involving blood-derived macrophages and central nervous system (CNS)-resident microglia. Intra-vital two-photon microscopy enables the study of macrophages and microglia in the spinal cord lesion in the living animal. This can be performed in adult animals with a traumatic injury to the dorsal column. Here, we describe methods for distinguishing macrophages from microglia in the CNS using an irradiation bone marrow chimera to obtain animals in which only macrophages or microglia are labeled with a genetically encoded green fluorescent protein. We also describe a injury model that crushes the dorsal column of the spinal cord, thereby producing a simple, easily accessible, rectangular lesion that is easily visualized in an animal through a laminectomy. Furthermore, we will outline procedures to sequentially image the animals at the anatomical site of injury for the study of cellular interactions during the first few days to weeks after injury.
The inflammatory reaction to disease or injury in the central nervous system (CNS) is poorly understood, especially with regards to the interactions among immune and resident cells within the tissue. Investigations of these cellular interactions in the spinal cord are of particular interest in the living animal. The only easily accessible CNS white matter tract is in the dorsal columns of the spinal cord, making this an important area on which to focus efforts on improving feasible experimental approach. These investigations have been limited due to the technical difficulty in accessing and stabilizing the CNS for imaging. Live imaging in the spinal cord has been described previously1-13, however, few studies have addressed cellular movement in a spinal cord injury beyond a few hours after the initial insult. The traumatic spinal cord lesion is a complex environment, with neurons, astrocytes, fibroblasts, NG2 progenitor cells, and immune cells including microglia, neutrophils, macrophages, T cells, B cells and dendritic cells14,15. Macrophages are the subset of immune cells responsible for phagocytosis in the lesion, infiltrating from the circulation. The role of these phagocytic cells has been debated, with reports indicating that these cells can take on both damaging and protective roles in the injured tissue. These roles range from increasing secondary axonal dieback after an injury and acting in a phagocytic manner16-22, to taking on a wound healing phenotype and decreasing functional deficits in the injured animal21,23,24.
Previously, attempts to distinguish macrophages from microglia have relied primarily on morphology in healthy tissue. However, activated microglia and macrophages express many of the same markers and display indistinguishable morphology after injury25-29, making the separation of different activities of these cells difficult to study based on these factors alone30. These cells can be separated by differential expression of CD45 using flow cytometry33,34, although this approach is less useful in discriminating these cell types in vivo. Utilizing differential expression of CCR2 and CX3CR1 in microglia and macrophages has also been explored, although the dynamic changes in the expression of these markers as monocytes differentiate into macrophages can complicate accurate analysis35,36. Microglia are the resident immune cells in the CNS and arise from yolk sac progenitors during fetal development, while macrophages are derived from bone marrow progenitors and enter the injured CNS after an insult31,32. An alternative approach to using morphology to distinguish these two cell types is to replace the bone marrow with progenitors from a donor animal expressing a traceable marker in the macrophages derived from the donor progenitors, while preserving the CNS resident, recipient-derived microglia. These bone marrow chimeric models are commonly utilized in many other applications37-40. This method has unique caveats associated with damage to the integrity of the blood-brain barrier caused by irradiation or cytotoxic drugs used to eradicate marrow progenitors in the host, thereby limiting its use in certain applications. Recently, functional distinctions between microglia and macrophages in the CNS are beginning to be discerned by using flow cytometry, gene chip array analysis and chimerism methods24,26,41-44, which will prove very useful in the future. Although separating these two cell types remains difficult, understanding their unique functions is important in leading to more targeted treatments of disorders that involve both microglia and macrophages within the CNS.
Description of cellular interactions within the spinal cord lesion has primarily come from studies involving cell culture models. This is due to the challenges involved with imaging both the spinal cord lesion and immune cells together in a living animal. Current rodent models of spinal cord injury of varying type and severity include contusion models45,46, pin-prick injuries2, laceration47, and the dorsal column crush injury described here16,18,48. Inflammation in the meninges increases with the severity of injury and poses unique challenges for the implantation of windows or surgeries for serial imaging. Some of these challenges include the infiltration of phagocytic cells as well as the generation of fibrous tissue over the surgical site. Some of these issues can be overcome by creating smaller lesions and covering the exposed spinal cord with non-immunogenic surgical dressing between the dura and the paraspinous muscles to allow for subsequent re-opening surgery, as is done here to study the dorsal column crush injury. The ability to image only the dorsal portion of the spinal cord also limits the choice of injury, since contusion models primarily cause damage to the central cord, often sparing the most dorsal part of the dorsal columns, especially at early time points after injury45,49. Therefore, we describe a simple injury model that yields a clean, repeatable, rectangular shaped lesion that is useful for both the observation of cell movement within the lesion and for easy quantification of axonal position relative to the lesion.
NOTE: All animals should be housed and utilized in accordance with the animal care and use committee (IACUC) at the institution. All procedures described here are approved by the Case Western Reserve University IACUC.
1. Transgenic Animals
2. Bone Marrow Chimeras
3. Dorsal Column Crush Injury
NOTE: Choose appropriate chimera animals or double transgenic animals for surgery based on the experiment desired. Animals with either labeled resident or bone marrow derived cells were created in previous steps.
4. Intravital Imaging
A schematic diagram depicting the dorsal column crush lesion is shown in Figure 1A. After the lesion, the animal is prepared and stabilized for intravital microscopy using spinal clamps (Figure 1B). An image taken immediately after the dorsal column crush injury shows a clearly visible rectangular lesion with forceps tine insertion points that are clearly identifiable. Axons at the injury site are severed across the span of the entire dorsal column (Figure 1C). The integrity of blood vessels remains intact, and no evidence of vessel dye leakage was detected by fluorescence. To perform serial imaging of the same lesion over weeks, the muscles and skin can be sutured over the laminectomy for subsequent re-exposure. Similar lesion morphology is seen at later time points (Figures 1D, E). 5 days after injury, the forceps insertion sites are still visible but the lesion has begun to increase in size due to secondary injury caused by inflammation. The axons at the caudal end of the lesion have retracted from the initial site of the injury via a process called “axonal dieback”. The axons at the rostral end of the lesion exhibited Wallerian-like degeneration (Figure 1D, E). Between days 5 and 22 after injury, the size of the lesion has stabilized. Concurrently, a large influx of CX3CR1+ cells can be seen infiltrating in and around the crush lesion during these time points, although the identity of these cells (microglia versus macrophages) can not be distinguished in the preparation as shown in Figure 1.
In order to distinguish the behavior of microglia and macrophages within the crush lesion microenvironment, a radiation chimera model was used (outlined in Figure 2A). First, the bone marrow progenitor cells of a CX3CR1+/GFP mouse are replaced with non-fluorescent donor cells, thus only GFP+ CNS-resident microglia are visible by fluorescent imaging. The efficiency of marrow reconstitution can be confirmed by flow cytometry examination of the peripheral blood 8 weeks after marrow transplant (Figure 2B, C). In Figure 2B, F4/80 and GFP positive macrophages, highlighted in blue, are detected in a CX3CR1+/GFP → C57BL/6 chimeric mouse, in which some GFP negative but F4/80 positive monocytes are also present. In Figure 2C, no double positive F4/80 and GFP positive cells are left after replacement of CX3CR1+/GFP recipient bone marrow with wild type bone marrow. Before injury, microglia are evenly distributed within the spinal cord, displaying a resting, ramified morphology (Figure 2D). 8 days after injury, only a few microglia can be detected in and around the lesion. These microglia display an amoeboid morphology (Figure 2E). Second, bone marrow progenitor cells in a non-fluorescent mouse are replaced by marrow cells from a CX3CR1+/GFP mouse, therefore rendering bone marrow derived CX3CR1+ monocytes / macrophages visible by GFP fluorescence (Figure 2A). Before injury, the only GFP+ cells detected are largely perivascular, which have been reported to be bone marrow derived and continually turn over on a regular basis28 (Figure 2F). After a crush injury, CX3CR1+/GFP cells can be seen along the inside of blood vessels surrounding the lesion (Figure 2H-J), and the lesion center is filled with infiltrating CX3CR1+/GFP macrophages (Figure 2G).
Time lapse imaging of the dorsal columns can be performed for up to 6 hr. In Figure 3, we show a non-chimeric CX3CR1+/GFP mouse immediately after injury in which cells from the circulation can be seen moving out of the blood vessel towards the lesion core and moving within the injury site (Figure 3A, B; Supplemental Movie 1). Image analysis utilizing fluorescence intensity to identify cells can track the paths taken by macrophages (Figure 3A, B; Supplemental Movie 1). Statistics derived from the imaging analysis show the average motility of macrophages in the lesion is 3.6 µm/min (Figure 3D). Finally, at 22 days after injury, axons, microglia and macrophages can still be detected within the lesion demonstrating the area of imaging is stable over long periods of time (Supplementary Movie 2).
Figure 1: Creation and sequential imaging of the dorsal column crush injury. (A) The dorsal column crush injury is performed by removing the dorsal process of the vertebra at the level of interest. Forceps are inserted into the dorsal column of the spinal cord 1 mm apart and are then closed 3 times to produce the lesion shown in purple. (B) The animal is then positioned with spinal cord clamps to obtain stability for intravital imaging. (C) Immediately after injury, the forceps insertion sites (red oval) and the rectangular crush injury volume (grey box) can be clearly identified. Axons in the dorsal roots can be seen (white arrows), as well as the injured ends of the axons on the caudal side of the lesion (filled arrow heads). (D) 5 days after injury, the forceps insertion sites (red oval), and the extent of the lesion (grey box) can still be easily visualized. The lesion has increased in size since the initial insult. Axons undergoing Wallerian-like degeneration can be seen rostral to the lesion (open arrow heads) in addition to axons in the dorsal roots (white arrows) and the ends of injured axons (white arrowheads). (E) 22 days after injury, the lesion site is still identifiable with similar dimensions to the injury after 5 days. Note the large number of CX3CR1+ cells in and around the injured site. Features labeled are the same as in (D). All scale bars = 200 μm. Please click here to view a larger version of this figure.
Figure 2: Construction of chimeric mice to track CX3CR1+/GFP expressing CNS-resident microglia or bone marrow-derived macrophages in the dorsal column crush lesion. (A) A schematic diagram of chimeric mice generation with either CX3CR1+/GFP expressing microglia or macrophages. First, Thy-1 YFP H mice are irradiated and the bone marrow is reconstituted with marrow cells isolated from a CX3CR1+/GFP mouse, resulting in a chimeric mouse whose macrophages are GFP+. Second, Thy-1 YFP H / CX3CR1+/GFP mice are irradiated and the bone marrow is reconstituted with marrow cells isolated from a non-fluorescent mouse, producing a chimeric mouse whose microglia are GFP+. (B) Example of flow cytometry data with F4/80+ and CX3CR1 GFP+ cells in the blood from a chimeric animal with a CX3CR1+/GFP marrow donor transplanted into an irradiated C57BL/6 recipient. Cells derived from the CX3CR1 positive donors that are positive for both GFP and F4/80 are shown in the blue highlighted area. (C) Example of flow cytometry data with F4/80+ and CX3CR1 GFP+ cells in a chimeric animal with a C57BL6 donor marrow transplanted into an irradiated CX3CR1+/GFP recipient. Note the lack of double positive CX3CR1+/GFP and F4/80 cells within the blue highlighted area. (D) An intravital two-photon microscopic snapshot of a mouse from the first chimeric scheme, showing GFP+ microglia with ramified morphology within the dorsal column (arrow head). These cells are interspersed among axons (yellow) in the dorsal root and the dorsal column. Scale bar = 100 µm. (E) Intravital imaging of the same mouse in (B) 8 days after dorsal column injury reveals a few CX3CR1+ microglia with large and amoeboid shaped cell bodies that are lacking processes (arrow head). Scale bar = 100 µm. (F) Snapshot of a mouse from the second chimeric scheme in (A), showing scant GFP+ macrophages, within the CNS parenchyma. Scale bar = 100 µm. (G) Intravital imaging of the same mouse in (F) 8 days after dorsal column injury reveals a large influx of macrophages in and around the lesion. Scale bar = 100 µm. (H) A higher magnification picture of CX3CR1+ blood derived cells in contact with the vessels in the uninjured animal. Scale bar = 30 µm. (I) A reconstruction of CX3CR1 cells in contact with the vessel, viewed from inside the vessel. Scale bar = 30 µm. (J) A reconstruction of CX3CR1 cells in contact with the vessel, viewed from outside of the vessel. Scale bar = 20 µm. Please click here to view a larger version of this figure.
Figure 3:. Representative cell tracking data in a non-chimeric CX3CR1+/GFP mouse immediately after dorsal column crush injury. (A) A snapshot of CX3CR1+ microglia and macrophages around damaged axons (yellow) from Supplementary Movie 1 immediately after a dorsal column crush injury. (B) The paths (grey) taken by the CX3CR1+ cells in (A) during a 110 min intravital imaging session. (C) A time coded representation of the tracks in (B). Tracks are colored based on their time within the whole 110 minute movie, as shown in the time bar. (D) Histogram of overall motility of the CX3CR1+ cells. All scale bars are 30 µm. Please click here to view a larger version of this figure.
Supplementary Movie 1: Representative intravital movie immediately after dorsal column crush injury in the double heterozygote Thy-1YFP/+ / CX3CR1GFP/+ mouse. Intravital spinal imaging was performed immediately after a dorsal column crush injury at T11 level. Right panel shows the movement of CX3CR1+ microglia and macrophages. The paths of the cell movement are shown as white tracks on the left panel. The injury is located to the upper left corner. CX3CR1+ cells can be seen moving into the site of the injury along these tracks. Axons are shown in yellow. Scale bar = 40 µm. Total time: 110 min. Playback speed: 300x.
Supplementary Movie 2: Representative intravital movie at 22 days after dorsal column crush lesion in the double heterozygote Thy-1YFP/+ / CX3CR1GFP/+ mouse. Shown here is a representative movie taken in a mouse 22 days after the initial dorsal column crush injury at spinal level T11. The injury is located at the upper right corner of the frame. Axons (yellow) ascending rostrally are shown at the bottom right. The dorsal vein and blood vessels are labeled with TRITC-dextran (red). CX3CR1+ cells within the lesion move at a slower speed compared to those immediately after injury. Scale bar = 50 µm. Total time: 90 min. Playback speed: 600x.
Imaging interactions of different cell types in their native tissue compartments during ensuing pathology in real-time has generated great interest. Within the dense network of the CNS, cell-cell contact and signaling with adjacent cells are essential for normal function and for understanding CNS pathology. Here, we described the use of 2-photon laser scanning microscopy for the observation of cellular movement within a mechanical lesion in the spinal cord. In addition to the quality of the surgery and tissue preparation, mechanical stability of the tissue is paramount for successful time-lapse microscopy, particularly the isolation of the spinal cord from breathing motion artifact. Stability can be assessed by looking for movement of the spinal cord under a dissecting microscope that corresponds to the heart rate or breathing of the animal. Stability should be assessed both while performing the surgery and also immediately before beginning imaging. If the animal is not stable, adjustments should be made to the placement and tightness of the spinal cord clamps. Cell-type specific fluorescent reporter mouse models coupled with bone marrow chimera approaches have allowed the identification of monocytic cells versus microglia and axons. Previous studies have revealed that blood derived monocytes and not microglia are responsible for secondary axonal damage after trauma using the methodology described here43. Dextran conjugated vessel dye allows for vessel identification both to provide landmarks and to identify breaches in vessel integrity. The selection of the appropriate fluorescent reporter mouse model is critical to allow the proper identification of the desired cell types to be imaged.
Development of fluorescent mouse models that are appropriate for the studies being undertaken is also critical to the success of experiments. Distinguishing between the two phagocytic populations of CX3CR1+/GFP cells in the CNS has traditionally been difficult. As shown here, a common immunological technique, irradiation chimeras, can be utilized to distinguish radio-resistant, CNS-resident microglia from bone marrow-derived macrophages. The irradiation procedure has the potential to damage the blood brain barrier and alter cell phenotypes, so use of this model should be considered carefully. The extent of these changes differs with irradiation dose as well as duration of recovery period, and their impact on different inflammatory models has not been fully studied. The effects of irradiation on CNS blood brain barrier can be minimized by shielding the head during irradiation, and this has been shown to decrease cell infiltration into the spinal cord after irradiation, even if the spinal cord is not directly shielded52. Here we have observed that the number of cells infiltrating into the CNS at steady state as a result of chimera generation is insignificant in contrast to the number of cells entering the lesion. Other alternative models to consider include drug-induced chimeras52 or parabiotic mouse models53.
Here, we have presented a specific, simple and reproducible small injury model that results in a lesion to the dorsal white matter of the spinal cord that is easy to image using the 2-photon microscopy. This method also provides a quantifiable lesion to assay the degree of axonal dieback in the dorsal columns that in the literature has been coupled with complementary fixed tissue analysis16,18,43,48,54. In this model, animals display only minimal deficits and require no special care after injury. Future studies utilizing the dorsal column crush injury and imaging techniques described here may be powerful screening tools to assess the efficacy of treatment for spinal cord injury. These treatments may include small molecule inhibitors, drugs, cell products, tissue grafts and combinatorial treatments. The interplay between macrophages, microglia and neurons are also likely to play a role in other disease models in the spinal cord, including multiple sclerosis, tumors, meningitis and amyotrophic lateral sclerosis, and this technique may be useful in the study of these diseases as well.
The authors have nothing to disclose.
The following agencies provided critical funding support for this study: MSTP T32 GM007250 (T.A.E.), 5T32EB7509 (D.S.B.), NCI R01 CA154656 (A.Y.H.), Dana Foundation (A.Y.H.), St. Baldrick’s Foundation (A.Y.H.), Alex’s Lemonade Stand Foundation (A.Y.H.), Gabrielle’s Angel Foundation (A.Y.H.) and Hyundai Hope-on-Wheels Program (A.Y.H.). The authors are thankful for the indispensable help of Jerry Silver and Sarah Busch in learning the Dorsal Column Crush (DCC) injury. The authors are also grateful to Jingquang You, Elisabeth Hare and Hongmei Hu for technical help with genotyping and Ross Anderson for mounting the spinal stabilizing unit.
Name | Company | Catalog Number | Comments |
CX3CR1 GFP mice | Jackson laboratories | 5582 | |
Thy-1 YFP H mice | Jackson laboratories | 3782 | |
Mouse pie cage | Braintree Scientific | MPC 2 set | |
60mm2 petri dish | Corning | 430166 | For bone marrow isolation |
Falcon 40 um cell filter | Fisher Scientific | 08-771-1 | For bone marrow isolation |
1x15mL and 1x50mL conical tube | Fisher Scientific | For bone marrow isolation | |
16 gauge needle | BD | 305177 | For bone marrow isolation |
1ml syringe | BD | 309628 | For bone marrow isolation |
ACK lysis buffer | Gibco | A10492-01 | For bone marrow isolation |
HBSS | Gibco | 1E+07 | For bone marrow isolation |
RPMI media | Sigma | R8758 | For bone marrow isolation |
F4/80 antibody | Ebioscience | 12-4801 PE | For FACS verification of chimeras |
30-gauge insulin syringe | BD | 328280 | For tail vein injection |
Spinal Cord Clamps | Narshinge | STS-A | For imaging |
Ortho-Jet Dental Acrylic powder | Lang Dental | REF1320 | For imaging |
Ortho-Jet Dental Acrylic liquid | Lang Dental | REF1404 | For imaging |
Gelfoam gelatin sponges | Pfizer | 9E+06 | For imaging |
4-0 Ethilon sutures | Ethilon | 1667G | For surgery |
Reflex Clips, 7mm | Kent Scientific | INS750344 | For surgery. Sterilize before use |
Iris Scissors | Fine Science Tools | 14061-09 | For surgery |
45/5 Forceps, Dumoxel | Fine Science Tools | 11251-35 | For surgery |
Rongeurs | Fine Science Tools | 16021-14 | For surgery |
30 gauge needle | BD | 305106 | For making access holes in dura |
Forceps Dumont #4 Biology tips | Dumont | 11242-40 | For surgery |
Needle Drivers | Fine Science Tools | 12002-12 | For surgery |
Michel Wound Clip Forceps | Kent Scientific | INS700753 | For surgery |
Wound clip remover | Fine Science Tools | 12033-00 | For surgery |
O-rings for Forceps | Fine Science Tools | 11200-00 | For surgery, Often provided with #4 forceps, can also be ordered separately |
Cordless Rechargable Animal trimmer | Wahl | Series 8900 | for surgery |
Vetbond | 3M | 1469SB | For surgery |
Betadine Solution (10% Providine-Iodine Topical Solution) | Purdue Products L.P. | NDC 67618-150-08 | For surgery |
Nair Hair remover lotion | Church & Dwight Co., Inc. | NRSL-22329-05 | For surgery |
Isoflurane | Butler Schein | NDC 11695-6776-2 | Drugs – for surgery |
Marcaine | Henry Schein | 6E+06 | Drugs – for surgery 1.0 mg/kg given subcutaneously |
Bupernex | Henry Schein | 121-7793 | Drugs – for surgery buprenorphine 0.1 mg/kg |
aCSF | Chemicals from Sigma | 119 mm NaCl | For surgery |
26.2 mM NaHCO3 | From Cold Spring Harbor Protocols | ||
2.5 mM KCl | |||
1 mM NaH2PO4 | |||
1.3 mM MgCl2 | |||
10 mM glucose | |||
TRITC-dextran 150,000 MW | Sigma | T1287 | For intravenous administration to label vasculature |