I am not convinced that you can use morphology *alone* to claim that microglia can be conclusively categorised as a specific phenotype (and that's presuming the phenotype names that we use are accurate descriptions of the status of individual microglia...).

Loss of processes and adoption of a rounded/amoeboid shape is widely accepted as M1, but I don't think any morphologies have been clearly shown to be identifiable as any M2 variant (happy to be corrected here).

There are numerous papers describing microglial morphologies (in 3D/tissue), and correlating them with changes in gene/protein expression typical of pro-inflammatory (M1) and 'alternative' activation (M2a/b/c) status. You should start with this literature.

If you can gather qPCR, Western blot and morphological data, perhaps you can answer this question yourself. I haven't seen it conclusively demonstrated elsewhere, but I'm new to this field, and very happy for someone to tell us where to look.

If you show that specific morphological features are ~always associated with accepted pro- or anti-inflammatory gene/protein changes, then you will have demonstrated that specific morphological features can be reliably assumed to indicate specific microglial phenotypes.

Final comment: microglia have such active membranes/cytoskeletons that no single shape is likely to serve as an accurate archetype/template for each phenotype. They are constantly remodelling themselves, and in time-lapse/dynamic imaging studies, a single cell may appear amoeboid in 1 frame, and within 1 minute could extend numerous processes, appearing ramified. I suspect this will prevent morphology alone ever being enough to categorise individual cells; although you could probably categorise a microglial population's 'average/typical phenotype' based on 'average' morphology.

Dear Bhakta,

I think that activated microglia has significantly less and shorter  branches or processes (amoeboid shape). In principle, it is believed that on activation, microglial cells transform from a ramified resting state into a reactive amoeboid state. This is believed to lead to complex changes in their physiologic and morphologic properties, with a substantial increase in phagocytic capacity.

For an example of activated microglia morphology please see Figure 9c, 9d in my following publication. Microglia in this experiments were activated by pharmacological blockade of GABAergic and Glycinergic inhibitory synaptic transmission.

Best wishes, Refik

Article Emerging Roles of Filopodia and Dendritic Spines in Motoneur...

I am not convinced that you can use morphology *alone* to claim that microglia can be conclusively categorised as a specific phenotype (and that's presuming the phenotype names that we use are accurate descriptions of the status of individual microglia...).

Loss of processes and adoption of a rounded/amoeboid shape is widely accepted as M1, but I don't think any morphologies have been clearly shown to be identifiable as any M2 variant (happy to be corrected here).

There are numerous papers describing microglial morphologies (in 3D/tissue), and correlating them with changes in gene/protein expression typical of pro-inflammatory (M1) and 'alternative' activation (M2a/b/c) status. You should start with this literature.

If you can gather qPCR, Western blot and morphological data, perhaps you can answer this question yourself. I haven't seen it conclusively demonstrated elsewhere, but I'm new to this field, and very happy for someone to tell us where to look.

If you show that specific morphological features are ~always associated with accepted pro- or anti-inflammatory gene/protein changes, then you will have demonstrated that specific morphological features can be reliably assumed to indicate specific microglial phenotypes.

Final comment: microglia have such active membranes/cytoskeletons that no single shape is likely to serve as an accurate archetype/template for each phenotype. They are constantly remodelling themselves, and in time-lapse/dynamic imaging studies, a single cell may appear amoeboid in 1 frame, and within 1 minute could extend numerous processes, appearing ramified. I suspect this will prevent morphology alone ever being enough to categorise individual cells; although you could probably categorise a microglial population's 'average/typical phenotype' based on 'average' morphology.

Are you studying tissue sections? How thick?

From perfused animals? How long after ischemia?

yes I am using mouse brain section of 20 micrometer thickness. Mice were perfusion fixed with ice cold pbs followed by 4% PFA in 0.2M PB. brains were cryprotected in 30% sucrose and sectioned. I am doing 3 days after the ischemic insult.

Hi we measured microglial activation in previous papers (please see below). We measured 36h after the insult. Microglial activation is a transient phenomenon. Right timing is very important for this phenotype. 

Journal of Neuroscience. 2016, DOI: 10.1523/JNEUROSCI.0426-16.2016, PMID: 27277809

Antioxidants & redox signaling. 2013, DOI: 10.1089/ars.2011.4491, PMID: 22746536

Please have a look at this paper:

Zanier ER, Fumagalli S, Perego C, Pischiutta F, De Simoni MG. Shape descriptors of the "never resting" microglia in three different acute brain injury models in mice. Intensive Care Medicine Experimental 2015; 3(1): 39.

Activated microglia  has a short processes and bigger soma - Please have a look on this paper  J. Neurosci. 2004 24 (39) 8500-9

good luck

Morphology does not tell microglial functions. Amoeboid morphology suggest that microglia is "activated", possible phagocytic. However, I found that microglia with perfectly ramified morphology can be opsonizing amyloid beta; Fig 9 in my J.Neuroinflammation paper http://jneuroinflammation.biomedcentral.com/articles/10.1186/1742-2094-8-152

Our recent live cell imaging has demonstrated exactly what SI Jenkins above stated; microglia changes morphology extremely fast and regardless its morphology can phagocytize neuronal processes. (Sorry no videos of figures, since this is still unpublished.) 

Each stimuli causes unique cytokine secretion profiles, so one can't just focus on one or two cytokines, but need to evaluate the cytokine data as a whole, i.e. pro-vs. anti-inflammatory and maybe even check the trophic factor levels. This is naturally quite difficult to do in a single cell, especially if you are trying to evaluate brain sections and microglia activation phenotype. That being said one can use different antibodies to detect microglia and clever selection of microglial markers, phenotype indicators potentially included with surface molecule expression might allow one to make some conclusions of their activation phenotype. Personally I like to use Iba1 and CD11b as a surface markers. Iba1 labels most of the brain microglia (but also macrophages) allowing one to do morphological evaluations. Expression intensity might increase upon injury/activation. CD11b (a.k.a. Mac1) (similar to Galectin-3, a.k.a. Mac-2) is heavily upregulated in injured brain areas (i.e. after stroke, TBI), but also upon inflammation (e.g. induced by LPS). The phenotype marker selection depends on injury and time of the detection. Elevated iNOS expression typically indicated M1 activation and there are several good anti-iNOS antibodies available. If you are more interested in microglia phenotype markers I recommend Dr. Carol A Colton's reviews on the topic.