In the present communication, we explored the curious morphology of rod microglia to expand on the limited understanding of their role in neuropathology. Here we take advantage of the reproducibility and large quantity of rod microglia produced in the cortex after experimental diffuse TBI. Rod microglial cells within the S1BF were most numerous at 7 days post-injury, where they align perpendicular to the dural surface post-injury. Based on the results from the CD68 and OX-6 staining, rod microglial cells are immunophenotypically distinct from other microglial morphologies. It had been expected to observe co-localization of rod cells with a subset of immune markers. This was the case with CD68; however, the discontinuous labeling of OX-6 suggests a further division of rods that warrants investigation. Trains of rod microglia were associated with neuronal elements rather than other glial cells. These data support reports dating back a century ago, which illustrated the existence of rod-like cells adjacent to neuronal processes. Moreover, our results add support to the more recent hypothesis that microglia play a critical role in the recovery from CNS injury (reviewed in
The appearance of microglia has been reported to reflect their function
[28, 29]. Under normal physiological conditions microglia have a small, somewhat elongated cell body with long, fine processes. These ramified microglia are rather evenly spaced throughout the brain, with their processes pervading the entire brain. It is generally accepted that ramified microglia constantly survey the CNS and synapses for intruders/stresses which may disrupt structure and function of neuronal circuits
[30, 31]. This is highlighted by the observation that areas with fewer synapses have fewer microglia, such as white matter
Various stimuli can rapidly activate morphological change in microglia. Cellular processes thicken and begin to retract from their surroundings
. At this stage these cells produce molecules with known immunological functions
. Further along the lines of activation, microglial cell bodies become more amoeboid and their processes thicken further and decline in number. These cells are presumed to be motile and secreting cytokines, with the ability to move to the site of injury/insult to reduce damage
[28, 29]. Migrating microglia typically show an asymmetrical morphology, some suggesting that they have a leading and a trailing edge
. These microglia differ from the ones described in the current communication as these rod cells appear symmetrical and are mainly found in the cerebral cortex. In cases where tissue damage is extensive, microglia become fully activated and at this point they are morphologically indistinguishable from infiltrating macrophages. Fully activated microglia have an amoeboid cell body and fully retracted processes. Functionally, these cells are thought to phagocytose debris/invading pathogens
[28, 29]. In focal TBI, active and amoeboid microglia are found throughout the contused or cavitated tissue and thought to promote tissue degeneration and contribute to repair
[8, 11, 13–15, 27, 34]. In stark contrast, after diffuse TBI we found no amoeboid microglia/macrophages; nor evidence for ingested myelin or synaptic debris. Activated and rod microglia more likely contribute to the neuropathology and repair of injured circuits after diffuse TBI then the scavenging of debris.
Although the precise mechanism and the role of rod microglia coupling and train formation are yet to be identified, these cells are always observed to be end-to-end and never perpendicular to each other. The signals required to send microglia towards a rod morphology rather than a more traditional activation state were beyond the scope of the current study. Rod cells could be formed by migrating or proliferating microglia; not withstanding these cells ultimately align to neuronal elements. Indeed, microglia have been reported to migrate towards and adhere to damaged but surviving neurons, which allows for efficient uptake of diffusible molecules
. It is therefore plausible to suggest that rod microglia align to either less injured neurons or axotomized axons in an effort to limit damage. Alternatively, trains of rod microglia could just as easily align to protect uninjured axons. Clearly, rod microglia align with dendrites and axons of neuronal elements as described by Nissl over a century ago
. Co-labeling for trains of rod microglia and neuropathology would clarify reasons and associations of this microglia phenotype. However, we are technically limited at this time. Silver degeneration staining is incompatible with Iba1 immunohistochemistry. Double-labeling with the amyloid precursor protein (APP) marker of axonal injury was not pursued, because the antigen presentation is transient after experimental TBI, being absorbed by 72 h post-injury
[12, 35]. Future studies on a comprehensive neuropathological time course will include APP staining at earlier time points post-injury.
In our model, diffuse brain injury leads to perisomatic axotomy and argyrophilic localized to the S1BF over 7 days post-injury
[12, 36]. The rod microglia morphology was most evident at day 7 post TBI, coinciding with the highest levels of neuropathology
; rod microglia may align with these damaged axons. Furthermore, rod microglial cells were observed concomitantly with reduced staining of CNPase indicating a loss of myelin detection. Cascades of neurodegenerative and regenerative events may hit a crossroad at day 7, which catalyzes the maximal coupling of rod microglia. We believe rod microglia are contributing to circuit reorganization in the wake of neuropathology; the detrimental and beneficial effects of which are yet to be determined.
Here, we demonstrate for the first time a preponderance of rod microglia following experimental diffuse TBI. At this time, it remains unclear whether higher or lower levels of injury would alter the number or duration of rod microglia. Previous studies have proposed that rod microglia only form in tissue which is devoid of cavitation and contusion
, suggesting that experimental models of focal TBI and stroke would result in an unfavorable environment for rod-microglia formation. Indeed, preliminary evidence indicated that rod microglia can be found in experimental blast-overpressure brain injury
Some could question the existence of rod microglia, and presume the formation of trains to be radial glial cells. We however argue that this is not the case due to distinct morphological (Figures
3) and phenotypic differences (Figure
4). Radial glial cells are described as a single cell with long processes
[22, 23]. Additionally, reports indicate that radial glia will stain with a variety of markers including GFAP, S100β, and vimentin (reviewed in
). The rod microglial trains in the current study are composed of multiple cells as evidenced by the nuclear staining and they failed to co-localize with radial glial markers (S100β and vimentin, images not shown). The cell morphology described here is not one for radial glia, but rather a distinct microglia morphology, with one or more phenotypes.
It is intriguing to contemplate a functional role for the trains of rod microglia. One possibility is synaptic support, whereas synaptic stripping is equally likely
. Indeed, complementary work from our laboratory highlights a functional role for the S1BF (predominant location of rod microglia) in behavioral morbidity
. The experimental model of midline fluid percussion offers the opportunity to examine the diffuse injured brain in the absence of contusion and cavitation. It has been described as a model of circuit disruption rather than destruction
[12, 20]. Disrupted circuits reorganize
[36, 39] and mediate late-onset sensory sensitivity, where brain-injured rats exhibit whisker sensitivity (whisker nuisance task) that develops over 28 days post TBI
. The circuitry used by rats in the whisker nuisance task incorporates the S1BF, the region we have documented to have large numbers of rod microglia by 7 days post-injury. Taken together, these data implicate rod microglia in the development of late-onset morbidities. However, more work is required to clarify the precise role that rod microglia play in behavioral morbidity.
Mechanisms of microglia activation and the functional implication on neuronal circuit function remain understudied. Microglia influence the injury magnitude and recovery in neurotrauma patients through numerous mechanisms, including clearing of cellular debris, neurotoxicity, promoting and directing axon growth, or mediating the reorganization of neuronal circuit structures
[40, 41]. The role of microglia after injury has yet to be fully elucidated. However, depending on the activation stimulus (for example, APP or IL-4) the phenotype of the microglia may be neuro-toxic or -trophic
. Additionally, a well-controlled helper T-cell response induces the microglia to adopt a phenotype that facilitates neuronal protection and tissue repair
[42–44]. Of note, resident microglia can play a role in tissue repair, similar to that described for resident macrophages in peripheral organs (reviewed in
). Furthermore, synergistic effects of microglia and astrocytes are needed for tissue remodeling after lesions, as they contribute to the re-establishment of the blood-brain barrier, as well as by secreting anti-inflammatory cytokines to suppress the inflammatory cascade