Characterizing microglia activation: a spatial statistics approach to maximize information extraction.

Davis BM, Salinas-Navarro M, Cordeiro MF, Moons L, De Groef L. · 2017

Read Characterizing microglia activation: a spatial statistics approach to maximize information extraction. by Davis BM, Salinas-Navarro M, Cordeiro MF, Moons L, De Groef L. (2017) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/europmc-pmc5431479.txt.

Introduction

Davis and co-authors build spatial statistics tools to characterize microglia activation from microscope images. Their methods study appeared in Scientific Reports. Microglia guard the central nervous system. They probe for distress and rush to injury sites. Their behavior can protect or harm neurons. Judging which role dominates is hard. Morphology labels lack any standard. Resting and activated are judged by eye. The team wanted unbiased numbers instead.

Their approach pairs image analysis with spatial statistics. It tracks cell shape and global density. Nearest neighbor distance adds spacing information. A regularity index captures pattern order. Cluster analysis uses soma size and roundness. That clustering split microglia into phenotype groups. The test case was a mouse optic nerve injury. New behavioral insights emerged from the phenotyping. The authors offer the tools as generic. Any brain region or disease state can use them. Fewer animals are needed per answer.

Catalog record

CREATORDavis BM, Salinas-Navarro M, Cordeiro MF, Moons L, De Groef L.
YEAR2017
COLLECTIONPAPERS
FORMATTXT
FILE SIZE53 KB
LICENSECC BY 4.0
RIGHTSCC BY
SOURCEEurope PMC

Source: https://europepmc.org/articles/PMC5431479

Browse the full library · About the Tower of Babel

Full text

The opening of the work, from the library's hosted text:

Characterizing microglia activation: a spatial statistics approach to maximize information extraction By Benjamin M. Davis, Manual Salinas-Navarro, M. Francesca Cordeiro, Lieve Moons, Lies De Groef Published in: Scientific Reports DOI: 10.1038/s41598-017-01747-8 ABSTRACT Microglia play an important role in the pathology of CNS disorders, however, there remains significant uncertainty about the neuroprotective/degenerative role of these cells due to a lack of techniques to adequately assess their complex behaviour in response to injury. Advancing microscopy techniques, transgenic lines and well-characterized molecular markers, have made histological assessment of microglia populations more accessible. However, there is a distinct lack of tools to adequately extract information from these images to fully characterise microglia behaviour. This, combined with growing economic pressures and the ethical need to minimise the use of laboratory animals, led us to develop tools to maximise the amount of information obtained. This study describes a novel approach, combining image analysis with spatial statistical techniques. In addition to monitoring morphological parameters and global changes in microglia density, nearest neighbour distance, and regularity index, we used cluster analyses based on changes in soma size and roundness to yield novel insights into the behaviour of different microglia phenotypes in a murine optic nerve injury model. These methods should be considered a generic tool to quantitatively assess microglia activation, to profile phenotypic changes into microglia subpopulations, and to map spatial distributions in virtually every CNS region and disease state. INTRODUCTION Whilst small in size, microglia are of enormous importance for CNS health and repair. Fulfilling their neuroinflammatory role, these so-called ‘resident macrophages’ continuously probe the CNS environment for signs of distress, and migrate to sites of injury/infection/neurodegeneration where they recruit and activate additional cells, proliferate, phagocytose, clear debris and reorganize the CNS parenchyma. This microglial reactivity is a complex multistage activation process, which broadly encompasses their transformation from a ramified to an amoeboid morphology by first retracting the microglia processes and then extending dynamic protrusions, followed by cellular locomotion. Microglia dynamics are, however, sometimes very subtle, and go beyond their immune surveillance function. Indeed, microglia are believed to represent an important nexus between neurological and immunological activity in the CNS, and, for example, are also essential for synaptic homeostasis and regulation of neurotransmitters1–3. Despite their central role in mounting a rapid, protective response against perturbations of CNS homeostasis, microglial activity has also been linked to various pathological conditions including neurodegenerative diseases. Indeed, while microglial reactivity may at first aid to the elimination of the source of distress and to the restoration of tissue homeostasis, persistent inflammatory stimuli (as present in e.g. Alzheimer’s, Parkinson’s diseases, amyotrophic lateral sclerosis) may ultimately overwhelm microglia and lead to a downward spiral of chronic inflammation that compromises neuronal survival4, 5. Altogether, in response to their ever-changing surroundings, microglia display a remarkable degree of phenotypic plasticity – both in terms of morphology, number and molecular markers – and may be skewed to either a neurotoxic or a neuroprotective role. Much remains to be investigated, however, in order to understand this dynamic process and to correlate microglia phenotypes and function. Historically, microglia phenotypes have been described by categorizing cells based on their morphological features. In general, highly ramified cells were designated ‘resting’ microglia and amoeboid cells were called ‘activated’, with a range of intermediate activation states in between (e.g., ‘intermediately activated’, ‘bipolar’, ‘rod-like’, ‘hypertrophied’, ‘bushy’). No morphological classification standard exists, however, which complicates comparison of data from different studies, and prevents objective quantification of pathological status and therapeutic effects. Nevertheless, it is well-known that microglia morphological features and function are tightly coupled3, 6, and the development of unbiased and quantitative methods could therefore lead to improved insights into this form-function paradigm. Furthermore, microglia reactivity is a highly dynamic process, which is likely to be better described by a continuum of quantitative physiological measures – able to detect subtle changes in cell morphology and to capture the dynamics of microglia plasticity – rather than by a qualitative categorization3, 7. Over the last years, novel methodological approaches for adequate description of the reactive transformation of microglia have been emerging (e.g., refs 3, 8–12), yet correlating microglia morphology to function remains difficult and few easy-accessible user-friendly tools are presently available to adequately quantify morphological changes. In addition, microglia phenotypes appear to heterogeneous, both amongst CNS regions and within a single tissue, which emphasizes the need for comprehensive analysis of the entire microglia population. In the current study, microglial reactivity was evaluated on whole-mounted mouse retinas, harvested four days after ONC. The ONC model has been extensively used to study retinal neurodegeneration, and the timing and pathological mechanisms leading to retinal ganglion cell death have been the subject of many studies, e.g. refs 13–16. Furthermore, a growing interest in the detailed course of microglial reactivity in this experimental glaucoma model has been emerging17–21. Aiming to address this recent focus, we designed an automated analysis algorithm that, based on Iba-1 (ionized calcium-binding adapter molecule 1) immunostaining of the retinal microglia, quantifies five parameters: cell density, nearest neighbour distance (NND) and regularity index, to describe microglia numbers and distribution; and cell soma size and roundness, to assess cell morphology. Together, in combination with a spatial statistics approach, this comprehensive set of measures allows us to extract greater information from whole-retinal histology than has previously been possible, to quantitatively describe the dynamic pathological state of microglia in the retina. RESULTS Quantification of microglial activation by a novel comprehensive image analysis script Retinal microglia were visualized by means of immunohistochemical staining for Iba-1, a commonly used macrophage/microglia-specific calcium binding protein and well-suited marker for morphological differentiation of ramified versus amoeboid microglia22. Automated detection and counting of Iba-1 immunopositive cells was validated by comparison to manual counts by three masked observers. Algorithm results correlated strongly with mean manual observations and subsequent Bland-Altman analysis23 revealed comparable inter-operator variability between manual observers and the algorithm described herein (Fig. 1, Table 1).Figure 1Algorithm optimization. (a) Linear fit of the mean operator counts versus the numbers of microglia detected by the automated algorithm, reveals a strong correlation (y = 1.007x; R2 = 0.843) and validates our algorithm. (b) This is confirmed by the Bland-Altman plot, which shows a bias of 4.20%. For comparison, bias for three observers was −8.35%, 4.33% and 3.50%. Data are depicted as mean ± 95% confidence intervals. Table 1Correlation analysis of automated detection versus manual counting of Iba-1 immunopositive cells.SlopeR2Pearson Rp-valueBias (%)95% CI Upper95% CI LowerObserver 11.0720.9200.9787<0.0001−8.3476.626−23.32Observer 20.9630.8730.9479<0.00014.32821.77−13.12Observer 30.9560.8850.9620<0.00013.499−14.0921.09Algorithm1.0070.8430.9376<0.00014.20319.61−28.01 Analysis of cell number and distribution to measure microglial reactivity The first set of parameters that was analysed in our data set comprised cell density, NND and regularity index, and aimed to assess microglia numbers and distribution. First, upon comparison of microglia cell numbers in retinas of naive mice versus retinas that underwent ONC (4 days post injury), overall cell density was found to dramatically increase by 43% (185 ± 5 cells/mm2 versus 264 ± 9 cells/mm2, respectively; P < 0.005) (Fig. 2a). This was also reflected in the NND, which decreased in the ONC retinas (40.8 ± 1.0 μm versus 34.2 ± 0.7 μm; P < 0.005) (Fig. 2b). A detailed analysis of microglia density and NND in function of the distance from the ONH, furthermore revealed that microglia density tends to be

Read the complete text