Morpho‐functional variation and response pattern of microglia through rodent ontogeny showing infant microglia as stable and adaptive than matured

Abstract Microglia, myelo‐monocytic lineage cells, that enter in the developing brain at early embryonic stages and integrate in CNS, are involved in almost all neuroinflammatory conditions. We studied how microglia change their responses through the development and maturation of brain in normal physiological conditions using an ex situ model to delineate their age‐specific morpho‐functional responsiveness. Rapidly isolated microglia from different age‐matched rats were characterized with Iba1+/CD11b/c+/MHCclassII+, cultured, studied for cell‐cycle/proliferative potency, ROS generation and phagocytosis, viability and morphological analysis induced with GMCSF, MCSF, IL‐4, IL‐6, IL‐10, and IFN‐γ. The study showed marked differences in cellular properties, stability, and viability of microglia through ontogeny with specific patterns in their studied functions which were coherent with their in situ morpho‐functional attributes. Phagocytic behavior showed a notable shift from ROS independence to dependence toward maturation. Perinatal microglia were found persistent in ex situ environment and neonatal microglia qualified as the most potent and versatile responders for morpho‐functional variations under cytokine induced conditions. The study identified that microglia from infants were the most stable, adaptive, and better responders, which can perform as an ex situ model system to study microglial biology.

Plenty of works are now revealing the variety of microglial functions related to different neurological disorders and other adversities. The changes in morpho-functional aspects from development to growth and maturation and aging have drawn attention till date, and our understanding evolved from the traditional ramified and amoeboid concepts to contemporary resting/surveillent functional reality of microglia in living tissue (Del Rio-Hortega, 1932;Nimmerjahn et al., 2005;Streit et al., 1999;Wake et al., 2009). With the emerging opinion on developmental plasticity and cellular heterogeneity or adaptability of the so-called microglia or brain antigen presenting cells or monocyte progenitors of brain (Ghosh, 2010;Santambrogio et al., 2001), it is now assumed that insight to the changes in microglial behavior in a normal growth continuum or ontogeny may possess the key to the core understanding of microglial response in various maladies in brain from diseases to aging. Therefore, postulating that microglia change with the maturation continuum, we tried to develop an ex situ model system to work with microglia from different developmental and maturation age groups by adopting a rapid isolation protocol for the cells from the rodent brain. Analyzing some basic cellular properties for maturation, morphological response, stability, and adaptability or variability against an array of cytokines were documented to determine the best responding microglial age group. This was a challenging attempt to assess a delicate cell like microglia ex situ from different age groups and stabilize them in culture for further experimentation. This may provide us valuable insights of microglial biology in a normally developing and maturing brain, and offers an effective model system to work with microglia in a similar platform for more intricate aspects of its response behavior in a developing age continuum or ontogeny.

Animals and grouping
The female Sprague-Dawley rats were maintained for the experiment, and were fed with hind liver pellets or equivalent and water ad libitum; 12 h light and dark cycles were maintained, examined, and weighed at regular intervals throughout the experimental period. Reproductively matured male and receptive females were set for breeding at a rate of 1:2, respectively, examined for copulation plug and other symptoms for mating to separate pregnant mothers and pregnancy days were counted to obtain the required embryos. Neonates were maintained with their respective mothers in one cage as they were at waning age.
The groups of animals maintained were-a) Late embryo (ED 18 ± 2), b) neonate (D 5 ± 1), c) young adult (D 45 ± 5), and d) mature adult (D 240 ± 10), where ED = embryonic days (calculated from the day of mating/copulation) and D = days after birth. For late embryos, the pool of all embryos from each pregnant mother and for neonates, the pool of all pups from each waning mother, and each individual from the other two groups had been considered as n = 1. In each experimental setup, n = 5 repeats were done. Animals were not sexed for the late embryonic , neonate (D5), and young adult (D45) groups as it was not validly possible and for matured adult (D240 ± 10) the results of two males and three females have been accumulated to form a data set for statistical interpretation, hence neutralizing the sex effect.

Rapid isolation of microglia from different age groups
The rats were deeply anaesthetized with sodium pentobarbital (50 mg/kg body weight) and transcardially perfused with ice-cold PBS to ensure the removal of monocytic cells present in blood. Whole brain was dissected out and placed into ice-cold PBS containing 1% penicillin-streptomycin solution (P/S), mechanically dissociated and enzymatically digested for 30 min at 37 • C by 5-15 U type II collagenase (Sigma-Aldrich) and 500 U DNase I (Sigma-Aldrich). Cell suspensions in ice-cold PBS passed through a cell strainer (BD Falcon, USA) of porosity ∼70 μm for perinatal brains and ∼100 μm for adults to make a single cell suspension, allowed to adhere on a glass petri dish (DURAN, Czech Republic) for an hour in 5% CO 2 humified environment at 37 • C (CO 2 Incubator Galaxy 48S, New Brunswick, Germany).
The adherent cells were then recovered with 1X Trypsin-EDTA (Sigma-Aldrich) solution followed by addition of DMEM (Gibco, Life Technologies, USA) with 10% FBS (Gibco, Life technologies, USA). Recovered cells were resuspended in media, laid on 20%/70% Percoll (GE Life Sciences, USA) gradient and centrifuged at 2500 rpm for 25 min at 18 • C (Thermo Scientific-HERAEUS Megafuge 8R, USA). Collected cells with highly enriched microglia from the interface were suspended in DMEM containing 1% P/S antibiotic antimycotic solution (Himedia, India) along with 10% FBS. Their viability and density were immediately measured by the trypan blue exclusion method in Neubauer improved chamber (Marienfeld, Germany) and observed under Microscope Nikon Eclipse TS 100 (Nikon Corporation, Japan). In the case of embryos and neonates, the whole brain of 5-6 pups or embryos were pooled together and treated with half-dilution of enzyme concentration for the initial steps of brain tissue digestion in contrast to adult rats and with some minor modifications in cell strainer and centrifugation.

Ultra-microscopy of isolated cells (scanning electron microscopy)
The cells immediately isolated from brain tissue were fixed in 2.5% glutaraldehyde (Sigma-Aldrich) for 4 h at 4 • C followed by washing in PBS. The sample was gradually dehydrated in graded alcohol and finally brought to 100% acetone followed by critical point drying, spread on a 1 cm 2 grease-free glass slide placed on a metallic stub with conducting silver paint. The cells laid on the glass platforms were then coated with gold-palladium alloy of 100-200Å thickness in a diode sputtering system. Finally, the samples were observed in FEI Quanta-200-MK2 (30KV), USA scanning electron microscope.

Immunophenotyping of isolated microglia from different age groups
The cells were suspended in PBS (1-2 × 10 6 cells/ml), mildly fixed with 2% paraformaldehyde, washed in PBS, and stained with antibodies as selected. Before staining, cells were blocked in 5% FBS in PBS (blocking buffer) for 1 h to avoid nonspecific binding. The cells were then washed, initially tagged and incubated with anti-Iba1 mAb (Abcam) for 1 h, washed and incubated with fluorescenated 2 • Ab in the dark for 1 h. After washing, the cells was again incubated with fluorescence conjugated CD11b/c mAb (Biolegend) in the dark for 1 h, washed, recovered, and prepared for flow cytometric analysis, assessed in BD FACSVerse flow cytometer, USA in blue laser (wave length 488 nm) excitation in FITC channel for Alexa Flour 488 conjugated CD11b/c and in PE channel for PE/Texas Red conjugated Iba-1 and reading were analyzed and represented with BD FACSuit software for immunophenotyping of the cells.

2.5
Approaches to establish microglia primary culture from different age groups of brain Immediately isolated cells from different age groups were prepared and attempted for primary culture. Initially, the cells were counted for assessing viability and density by trypan blue and morphological variations among different groups were observed and documented by Microscope Nikon Eclipse-TS-100, using CCD Camera (DS-Fi2-U3) and NIS Elements BR software (Nikon Corporation, Japan). As per cell count data, the isolated cell density was adjusted to seed for culture at a concentration of 1 × 10 6 cells/ml and cells were plated in 35 mm, 6 well and 12 well cell culture plates (BD Falcon Multiwell, USA) to establish primary culture for short and long term. Using both RPMI and DMEM media, finally it was standardized to be maintained in DMEM containing high glucose (D-glucose), L-glutamine + 15% FBS + 2% P/S. The cultures were placed in a 5% CO 2 humidified closed environment at 37 • C in CO 2 Incubator. The media of the culture were routinely changed at 3 days interval and scheduled to maintain for 2 weeks and more for all groups. The cellular morphology, viability, and immunophenotyping were scheduled to assess with the interval of 3 days, 6 days, and so on to check the culture conditions and to decide the continuity of culture.

2.6
Development of variable in vitro cytokine environment for microglia of different age groups, viability, and morphometric assessment After attempting to establish the primary culture for each group, to create an in vitro reductionist differentiation environment, we charged the selected cytokines separately in specific doses with days' interval with five different cytokines important for monocytic cell lineage differentiation. The cytokines applied were recombinant GMCSF, MCSF, IL-4, IL-6, and IFNγ procured from PROSPEC, Israel. Isolated cells from each group were seeded in duplicate for each cytokine in triplicate sets of experiments separately in the 5% CO 2 humified environment at 37 • C. After 24 h of seeding of cells at an average conc. 1 × 10 6 cells, they were charged with cytokines (MCSF,GMCSF, at variable concentration of 1, 3, 9, 50, and 100 ng/ml in 96-well plate for MTT assay for checking cell viability and at the variable concentration of 1, 3, and 9 ng/ml in 35 mm, 12 well or 24 well cell culture plates for cell culture for morphometric analysis and incubated in the same CO 2 humified environment with induced cytokines. The cells in culture were analyzed for their morphometric parameters to see the effect of these cytokines and cell shape determination. Their responses to each cytokine separately were measured at different concentrations. This study design was aimed to select the optimal cytokine dosing as well as the maximal responsive or sensitive microglial age group that could be handled in in vitro conditions to develop a model culture system. In these sets of experiments three different variables, that is, cytokines (MCSF, GMCSF, IL-4, IL-6, and IFN-γ), concentration of cytokines (1, 3, 9, and 50 ng/ml) and different developing/maturation ages of microglia (ED18, D5, D45, and D240) were taken into account where experiments for each age groups were done at least thrice in duplicate sets. The cellular adaptive morphology of microglia from different developmental age groups in culture with trigger of different cytokines was separately measured under bright field and phase contrast microscopy and photomicrographs of the cells were taken in Nikon microscope as mentioned earlier. Three major cellular morphological parameters were considered for morphometric analysis, that is, (1) single cell area, (2) single cell length, and (3) cell projection length, and measurements were done after 48 h (2 days) in culture. After analyzing this plethora of data (see statistical analysis also), the desired age group of isolated microglia for their responsiveness and optimal cytokine concentration to work has been generalized.

2.7
Immunostaining of isolated and cultured microglia from different age groups Immunophenotyping of cells had been done with CD11b/c (Biolegend) and Iba1(Abcam), for their acceptability as microglial markers and for culture additionally MHC class II (Biolegend) and Acridin Orange (AO) (Sigma-Aldrich) were also included counterstained with DAPI (Himedia, India). Fixing with 2-4% paraformaldehyde, they were washed and permeabilized by 0.2% Triton X-100 and aspirating Triton, the cells were rinsed in PBS followed by blocking in 5% FBS in PBS (blocking buffer) for 1 h. After washing differently conditioned cells in culture plates were incubated with diluted anti-Iba1 mAb for PE/Texas Red/Alexa Fluor 647 conjugates. Photographs were taken using CCD camera (DS-Fi2-U3, Nikon, Japan), processed, and analyzed with NIS Element BR Software (Nikon, Japan) for studying the immunophenotype of differently conditioned cells.

2.8
Analysis of cell cycle and proliferative capacity of microglia at different age groups

2.9
Assessment of phagocytic potential of microglia from different age groups The phagocytic potential of microglia in cultured condition after isolated from different age groups were measured by AO both by fluorescence microscopy and flow cytometry. Initially from the stock solution of 1 mg/ml of AO (Sigma-Aldrich) the working solution of 2.5 μg/ml in PBS was applied to the cells and kept in 5% CO 2 environment at 37 • C followed by removal of excess AO. Then the plates were viewed under the microscope by using either ter Block N G-2A (Nikon, Japan) as AO staining is visible in both the filters in green and red channels for nucleus and lysosomal vesicles respectively, photographed, processed, and analyzed with NIS Element BR Software (Nikon, Japan).
For flow cytometric quantification of AO uptake, after preparing the single cell suspension, as stated earlier, they were treated with AO at a conc. of 2.5 μg/ml to the cells (1-2 × 10 6 cells/ml) suspension and incubated at room temperature in dark, washed to remove excess AO and resuspended for acquisition in flow cytometer. The cells were assessed in BD FACSVerse flow cytometer, USA with blue laser (wave length 488 nm) excitation in PerCP-Cy 5.5 channel for AO and readings were analyzed and represented with BD FACSuit software.

2.10
Assessment of reactive oxygen species in microglia from different age groups The reduced non-fluorescent H 2 DCFDA can be oxidized and converted into fluorescent 2′, 7′-dichlorofluorescein (DCF) by intracellular ROS.
In this protocol, we applied H 2 DCFDA to label the intracellular ROS and detected the DCF intensity by flow cytometry. Cells were harvested by trypsinization, centrifuged, and washed with PBS and dissolved the pellet in PBS at a density of 1-2 × 10 6 cells/ml and then DCFDA (Sigma-Aldrich) was added at the final concentration of 20 μM.

2.11
Immunohistochemistry of formalin-fixed-paraffin-embedded rat brain tissue sections Deeply anesthetized (sodium pentobarbital, 50 mg/kg body wt.) rats intracardially perfused with 4% paraformaldehyde in 0.1 M PBS and with 5% sucrose were dissected out for brain and the tissue samples were post-fixed in same solution and stored. After routine histological preparation and sectioning of brain tissue cortex with 15 μm thickness were preheated overnight, xylene washed, gradually rehydrated, washed in PBS and mildly fixed in 2% paraformaldehyde, and washed. Blocking was done with 5% FBS in humified chamber at ambient temperature, washed and incubated with 1 • antibody overnight (1:700 dilution), that is, anti-Iba1 and anti-MHC class II (Abcam), and Alexa-Fluor 488 conjugated anti-CD11b (Biolegend), then washed and incubated for 2 • antibody, that is, PE/Texas Red conjugated goat antibody (Abcam) (1:1000 dilution) for an hour and washed. For double staining, same processes were repeated and finally nuclear staining were done with DAPI (Abcam), semidried and mounted in DPX and observed under Eclips TS 100-F epi-fluorescence microscope using DS-Fi2-U3 CCD camera and NIS-BR software (Nikon Corporation, Japan) with fluorescent filters as described in sub-section 2.7.
Silver staining-gold toning were employed to detect microglia in situ (McCarter, 1940). Slides with tissue sections were treated with ammoniacal aqueous solution after initial deparaffinization and passed through 10% Globus' hydrobromic acid solution. Washed slides were put into 50% aqueous silver carbonate, passed through 5% formalin, washed and placed in 1% aquous gold chloride solution, fixed in 5% sodium thiosulfate solution, washed and dehydrated in graded ethanol, finally mounted in DPX and observed under the same microscopic system in bright-field and documented as mentioned earlier.

Organotypic brain slicing and in situ observation of microglia with NDPase staining
After deep anesthesia as mentioned above the rats were intracardially perfused with 5% sucrose in PBS and brains were removed, cut into small squares and placed to filter papers with super-glue gel properly using tweezers and immediately submersed in PBS-sucrose solution.
Brain pieces of cortex region attached with vibratome platform by glue-gel were coronally sliced in Leica VT 1200S vibratome for 200 μm sections, collected and introduced in 0.1 M PBS with 7.5% sucrose at 7.4 pH (PBSS) and immediately processed for NDPase histochemistry (Almolda et al., 2013). The brain slices were replaced in 0.1 M TMB and pre-heated at 37 • C. Incubation medium was prepared with substrate Inosine-5′ di-phosphate, Sigma in TBM, added with manganese chloride and lead nitrate solution and filtered. Organotypic slices were placed in incubation medium at 37 • C for 30 min with shaking, removed, washed in PBSS, exposed to ammonium sulphide solution briefly, washed and stabilized by brief use of 1% silver nitrate solution, washed and mounted in gelatin-coated slides, dried, and counterstained with 0.1% toluidine blue. Then sections were dehydrated in graded ethanol, exposed in 1-butanol for 1 min and submerged in xylene, mounted DPX, and observed under Nikon TS 100-F, and documented as mentioned earlier.

Statistical analysis
All the assays were performed at least thrice or more for each age group using isolated microglia in culture and in duplicate sets or as specifically mentioned in the methods.

Rapid isolation of microglia from developmental/maturation stages and characterization
Microglial cells were isolated from four selected age groups with the modified protocol as mentioned earlier. Mostly, comparatively smaller cells with rounded morphology were isolated from the late embryonic and neonatal rat brain, whereas cells with elongated morphology were found isolated from adult brains (details of structural varia-

3.2
Establishment of microglia primary culture through developmental/maturation stages and immunophenotyping with lineage and antigen presenting markers

3.2.1
Establishment of primary culture One of the primary challenges of this work was to establish the primary culture of microglia from different age groups. After modification and Microglia from late embryos and neonates can continue their healthy morphology up to 3 weeks and more with repeated media changes in every 72 h. However, after 2 weeks they started to show changes in morphological features and adherence properties ( Figure 2B).
F I G U R E 2 Establishment of primary culture of microglia, receptor expression and characterization. Isolated microglia in culture for 5-6 days from different age groups shows healthy cell populations from perinatal groups (ED18 ± 2 and D5 ± 1) in the (A). The cells are showing moderate confluence and mostly rounded with few elongated morphology in culture. Elongation increases in neonatal cells than late embryonic cell isolates in culture with higher confluency. In contrary, unhealthy cells with granularity and apoptotic features are found after 5 days from adult groups (D45 ± 5 and D240 ± 10) while maintaining the same conditions. In (B), microglial cells isolated from neonatal rats kept in culture for 3, 6, 10, and 15 days in vitro (DIV), immunostained for their nuclei with DAPI (blue, in case of 6 day culture) and their morphology were analyzed. Phase contrast microscopy images at 3 DIV show microglia of mostly amoeboid with ovoid shape with only a few cells show ramified bipolar morphology. At 6 DIV, microglia cells become more homogeneous with more cells presenting highly branched processes forming ramified morphology. At 10 DIV, cells exhibit distinct monolayer of highly differentiated microglia with dense branching and at 15 DIV, showing enlargement of highly proliferating microglia cells extending their protrusions occupying whole surface. Microglial cell's nuclei are shown with DAPI (blue) cultured for 6 day (100X and 400X). In CD11b/c and Iba1 counter stained with DAPI. As adult cells tend to discontinue at 5-6 DIV, the poor number of cells with CD11b/c + co-expressed with Iba1 + has been documented for young adults only.
In contrary, cultured cells expressed core signature markers CD11b/c and intracellular Iba1 which are found to be expressed in almost entire population in microglia isolated and cultured from neonatal and late embryonic rats after 5-6 DIV. To observe that whether microglia continue to hold their functional attribute of antigen presentation in vitro, MHC class II expression of them from different groups in 5-6 DIV had been assessed with Iba1 expression as microglial lineage identity and counterstained with DAPI. Almost every cell was found to express MHC class II as found in the merged microphotograph for microglia with similar poor cell numbers in adults ( Figure 2C).  ing ∼7% of microglia in that age group ( Figure 3A). When microglia from different groups in culture were compared for their Ki67 expression in immunofluorescence microscopy counterstained with DAPI, it was observed that a particular population among them were Ki67 + and their intensity showed a decrease in postnatal and adult groups than late-embryonic group. When % of Ki67 + cells in each group were quantified in flow cytometry, it was found that the mean values for perinatal and young adult microglia expressing Ki67 in cul-ture were within 25-27% of the total population. However, a remarkable half-fold decrease occurred for matured adult microglia, restricting it around 13% of the population. This half-fold decrease of the proliferating marker Ki67 in matured adult microglia was found corroborated with the decrease of G2-M microglia found in the cell cycle phase distribution studies with PI and meant the downfall of the proliferating potential of microglia with maturation and aging ( Figure 3B).

3.4
Evaluation of phagocytic potential and reactive oxygen species production of microglia ex situ through developmental/maturation continuum In case of matured/aged rat, red spectrum emission was observed throughout the cytoplasm and almost all cells were AO + in their cytoplasm in accordance with a very high reading for AO in FACS and highly corroborative with the FACS data represented graphically in Figure 4B.

ROS production
It was found that when the ROS positive microglia from late embryonic brain showed a mean of ∼50%, the mean of neonatal ROS producing microglial population decreased to ∼30% or one-third of the popula- were showing minimal capacity of mean ROS production per cell, but it increased two folds in young adults from perinatal microglia. However, a remarkable hike in ROS generation potential occurred in cells which were isolated from matured aging brain and the observed increase was about 10 folds in comparison to perinatal microglia and about 5 folds from young adult microglia ( Figure 4C).

3.5
In situ morpho-functional attributes of microglia through developmental and maturing stages 3.5.1 Iba1 ± CD11b/c ± or MHC class II ± microglia in situ Immunohistochemical staining for microglia with cell specific markers showed the presence of microglia in the cortex region of the brain of different maturation ages. It has been observed that for late embryo Iba1 + CD11b/c + microglia are arrayed in tandem in many places of developing brain cortex where Iba1 + MHCclassII + microglia are dispersed. Presence of Iba1 + CD11b/c + and Iba1 + MHCclassII + cells increased in mature adult stage where higher expression of MHCclassII+ indicated toward increased antigen presenting function of microglia during mature adult stages ( Figure 5A).

3.5.2
In situ morphological variations Specialized staining technique with silver impregnation and gold toning in brain tissue classically applied to identify microglia in brain sections indicated on its distribution and morpho-patterning with ages. It is evident from microphotographs that their shapes were more rounded and condensed in late embryonic (ED16-18) and neonatal brains (D5) which transformed with branching projections in young adults (D45), whereas, considerable proportions of such cells showed an amoeboid morphology in mature adults (D240-270). This patterning was more prominently deciphered when NDPase reactivity staining was applied for observing microglia in different developmental/maturating stages ( Figure 5B). Rounded dense cellular cytoplasm of late embryonic (ED16-18) and neonatal (D5) stages transformed to highly ramified forms with projecting filapodia or fine branching surveilling brain parenchyma were observed in organotypic brain slices in the young adults (D45). Whereas, both ramified and amoeboid morphs of microglia were observed in mature adult brain slice with distinctive cytoplasmic functional activation with NDPase action ( Figure 5B).

Reactive microgliosis
The  Figure 5C). Such phenomena, classically termed as 'reactive microgliosis' had been documented in mature adult representing their reactive state in those age-matched brain. Such microgliosis, which more frequently observed in pathogenic brain tissue, was also documented in the normal aging brain indicating toward the presence of more active state of microglia, hence supporting our ex situ findings of phagocytic potential, ROS production, and related instability in primary culture.

Effect of different cytokines on microglia ex situ at different maturation stages on their viability and morphology
3.6.1 MTT assay for viability The results when compared, it was found that the viability of cells from the prenatal embryonic days was less (ranging within 0.04-0.13) com- groups initially (after 24 h of seeding), although they were unable to continue in culture for longer. Thus, within low viability ranges in general, microglia isolated from neonatal brain had shown better survival with a lower fluctuation range (within two folds) and had the potential to continue for a longer period in culture as previously seen ( Figure 6).

Choice of best dosing of selected cytokines
The best dosing selection for each cytokine for each group had been selected in short-term culture where the best responses were documented as the highest cell area, cell length, and/or projection length.

Selection of best responding microglial group in reductionist cytokine environment
As per the best responsive doses as selected from the above data sets, the cellular morphological parameters of each developmental and maturation age group were compared to determine which group is most responsive to the selected dose. The best dosing for every cytokine for each parameter, for each group were compared further to detect which F I G U R E 5 In situ morpho-functional attributes of microglia with development and maturity. In (A) distribution of microglia in brain cortex of different age matched groups, viz., late embryo (ED16-18), neonates (D5), young adult (D45), and mature adult (D240-270) have been detected in FFPE sections using microglia specific markers Iba1+ and CD11b/c+ in the first panel and in the second panel antigen presenting properties of distributed microglia in brain have been shown with Iba1+ and MHCclassII+ specific markers (Scale bar = 50 μm). In (B), distribution and morphological variations of microglia have been shown in different maturation states in first panel using silver-gold staining technique where rounded/amoeboid and ramified forms are indicated with red and blue arrows respectively (Scale bar = 10 μm). In second panel, microglial forms and ramifications are clearly visible and indicated separately by NDPase staining in organotypic brain slice. Presence of amoeboid cells in ED16-18 and D5 group transforms to ramified forms in D45 and redifferentiated to amoeboid shape in D240-270 (Scale bar = 50 μm). In (C), transformation of ramified microglia to amoeboid form in organotypic brain slice in the matured adult (D240-270) stage has been shown through steps 1 → 3 using NDPase staining (see text for details) group showed the maximum changes and variability in morphometry showed the highest variability in cell projection length when induced with GMCSF. Cytokines IL-4, IL-6, and IFN-γ showed maximized cell F I G U R E 6 MTT assay for detection of cellular viability of isolated microglia with different cytokines with variable concentrations. MTT assay to measure viability of microglia in culture in conditioned media by selected cytokines MCSF, GMCSF, IL-4, IL-6, and IFNγ respectively as shown in graph legends with variable concentrations from 1 to 100 ng/ml for each developing/maturing age groups where X axis designates cytokine concentrations in log scale and Y axis designates absorbance reading. The box plots are derived in Sigma Plot 13.0 software. Results differ among groups, cytokines and dosing (p ≤ .05), where ED16-18 group shows high viability, D5 group shows moderate viability and variability when exposed to different cytokines and dosing, D45 group shows low viability and high variability, and D240-270 group shows lowest viability and less variability. For details see text projection length for microglia in ED16-18 but almost marginally differed for IL-4 and IL-6 in microglia at D5, except for IFN-γ which varied considerably with maximized length for the age group ED16-18 in comparison to D5. From the above observations, we found neonatal infants and late embryos as groups which showed maximal changes in cytokine responses. However, most of the maximal changes shown by microglia in vitro were from neonatal (D5), that is, with 6 highest and 8 next highest median values out of 15 combinations (Figure 7).
There are different approaches to isolate microglia and continue with primary culture, but most have their limitations. Generally, two basic approaches are followed for microglia isolation from brain tissue.
One follows the method of prolonged shaking after preparing the mixed glial culture from tissue and other utilizes the density gradient centrifugation mostly from mechanically and enzymatically dissociated tissues (Frank et al., 2006;Ghosh et al., 2007;Hassan et al., 1991;Sedgwick et al., 1991). Other approaches are selecting and sorting cells by their receptors (Garcia et al., 2014;Gordon et al., 2011). For the first one, primarily prepared mixed glial suspension needs prolonged culture and shaking and the final yield of microglia requires weeks, thus altering the cellular properties from actual in situ condition. Alternatively, the enzymatically dissociated brain preparation is subjected to density gradient centrifugation for nearly an hour, causes stress to microglia with less yield. Moreover receptor-based sorting methods are rapid, but too restricted and dependent on expression of surface receptors that actually varies with development and age, hence causing partial recovery of microglial population in varied conditions.
Keeping these issues in mind and as per the demand of the present study, we approached by utilizing cellular properties of adherence with a short enzyme treatment following density gradient centrifugation (Ghosh et al., 2007;Hassan et al., 1991;Sedgwick et al., 1991). The purpose was to get cells rapidly from tissue and generate more yield in better functional conditions. Matrix adherence is a classic property of monocyte/macrophage cells (Giulian & Baker, 1986;Ling et al., 1983;Rustenhoven et al., 2016 ) and used here for a short period in comparison to prolonged adherence and incubation of mixed glial cultures in other methods (Giulian & Baker, 1986), thus helping in shortening centrifugation time and stress. These cells when characterized, showed required microglial properties ( Figure 1). Extending them in culture, healthy growth of microglia from perinatal stage was observed in contrast to poor continuation of cells in vitro from adult ( Figure 2).
Long term culture potential of perinatal microglia was also depicted in the study and extended beyond 3 weeks with almost cent percent confluency leading to a supra-basal layer formation of cells with differing morphotypes (data not shown). On the other hand, although we found exceptional purity of microglia from adults by immunophenotyping with Iba1+, CD11b/c+, and MHC class II+ marker expression (≥∼98%) in isolated cells, but showed poor response in culture and unable to extend in vitro. This stressed and restricted response of matured microglia ex situ can be correlated with their cellular feebleness found in functional parameters as observed later in this study, but also partially resulted from probable cellular damage during isolation due to their vulnerable extended and ramified morphology entangling the brain tissue in situ (Nimmerjahn et al., 2005;Rezaie & Male, 2002;Streit et al., 1999). This is indicating the reason that why already established most microglial cell lines from mice, rodent or human, namely, BV2 or N9, HAPI, and HMO6, respectively, are mostly developed from perinatal/neonatal brain and do not satisfy the age-oriented modelling of microglial function (Galatro et al., 2017;Horvath et al., 2008;Timmerman et al., 2018). This successful isolation of microglia from different age groups and establishment of primary culture for neonates with continuity fulfilled a challenging aspect of the study with further potential to work with the cellular model of microglia.
Proliferative and phagocytic potential both showed developmental and maturation age-sensitive variations (Figures 3 and 4). In general, microglia have been reported as having restricted proliferative capacity, but may alter varying degrees in differently conditioned brains (Askew et al., 2017;Gomez-Nicola et al., 2013;Mander et al., 2006).
However, there is no specific reporting for age-wise estimation of their proliferative capacity as we observed. Recent reporting of Askew et al., 2017 showed that microglia maintains a steady population by a spacio-temporal coupling of self-renewal process through a balance in between proliferation and apoptosis, and estimated this renewal process as pretty fast unlike previously thought (however much less than monocyte/macrophage) (Askew et al., 2017). Our observations showed specific trends in proliferative potential. Cells from matured adults showed a drastic reduction of G 2 -M phase, decreasing to nearly half of the previous groups, showing ∼6-7% of microglia in G 2 -M and similarly Ki67 showed microglial proliferation reduced half the amount in matured adults (∼13%) in comparison to other groups (∼24-27%) (Figure 3). However, the estimation of self-renewal ranging below 1% at a given time, as reported can be justified with the estimation of apoptotic spillage at that time in the neuropil, however such self-renewal may be faster if depleted in the natal brain (Askew et al., 2017;Huang et al., 2018). The strikingly low OD in microglia in MTT assay from the late embryonic brain indicates very less mortality and in reverse may have a chance to rapidly increase in number in the developing brains against a merely constant proliferating rate in all compared age groups. This may be due to a much higher difference between proliferation and mortality when average proliferation rates are not changing much within groups (Figures 3 and 6). Therefore, the prominent trend of decreased proliferating capacity with maturity leading to aging should be noted as an important performance indicator of the cell when considering the microglial physiology in any condition from maturity to aging. This observation can be correlated with another observation where alternatively activated microglia of M2 phenotype show retracted renewal capability and that might be the case in maturity and aging (Pepe et al., 2017).
Microglia is an obvious phagocytic mediator in the CNS for maintaining normal brain microenvironment or CNS organization to manage difficult neuropathological situations (Neumann et al., 2008;Sierra et al., 2013). Most of the studies were concentrated to find the role of phagocytic microglia in various neuropathogenesis until recently. But now its role in neuronal tissue and circuitry development are gaining attention with 'phagoptosis' and 'pruning' (Vilalta & Brown, 2018).
Microphotographs depicting microglia of late embryonic and matured age groups were the most efficient contenders of phagocytic behavior, which were also supported by FACS readings (Figure 4). It is well understood that microglia produce ROS in neuropathogenic inflammation due to degenerative damage or other stresses (Sierra et al., 2014;von Bernhardi et al., 2015). However, microglial ROS production when estimated through age continuum showed serious deviation with maturity.
The observed natural trend of remarkable many folds increase of ROS production in matured microglia may otherwise be one of the important causative factors of neuronal damage in maturation and aging.
Present study was aimed to decipher modifications of microglial vital functions through development and maturity and to identify their most stable and adaptive age-matched population which may serve as a suitable ex situ model for more detailed study on microglial cellular biology. Therefore, most of our studies were done on isolated microglial cells from brain parenchyma. Hence, simultaneous in situ observations were done to find the parity between their in situ states with ex situ behavioral manifestation. That, in turn, provided us clues to find connections and explanations for our observations and reduce artifactual redundancy. Increased MHC class II+ expressive microglia in mature adult stages with amoeboid transformation of microglia in that stage of maturity indicated higher phagocytic potential which was observed in our ex situ observation (Figures 4 and 5). Such transformation could be supported with the concept of 'reactive microgliosis' and high ROS production in aging microglia, and produced a probable cause of instability in primary culture ( Figure 2) (Rivest, 2009;Santiago et al., 2017;Streit et al., 1999). In contrary, ramifications with fine branching were found in young adult brain which were previously described as 'resting' but 'surveilling' microglia where we observed lower phagocytic potential with less ROS production (Nimmerjahn et al., 2005;Streit et al., 1999).
Such microglia with finely branched projections, we predicted, were nearly impossible to be retrieved during isolation and hence susceptible to damage, which might be a cause for lower stability and survival of them in primary culture ( Figure 2). However, neonatal and late embryonic microglia with their rounded but smaller morphological appearance in situ and less ROS productivity, higher proliferative potential ex situ were observed as stable microglial population is coherence with their properties.
Microglia, as an essential responder of brain microenvironment, are highly alert to survey the surroundings and modulate its activities accordingly (Nimmerjahn et al., 2005;Walker & Yirmiya, 2016). To operate with such sensitivity, they express large number of receptors and act as a source or sink of a plethora of soluble factors, cytokines, and chemokines (Hanisch, 2002). We hypothesized that such response behavior is changing with the developing and maturing ages. This was primarily observed with differential responses of sustenance and propagation in the primary cultures of microglia from different age groups ( Figure 2). Our investigation on microglia of different ages with selected cytokines, namely, GMCSF, MCSF, IL-4, IL-6, and IFNγ, which were already been reported to be involved in microglial differentiation (Ghosh, 2010;Hanisch, 2002;Kim & Nagai, 2010), showed varied degrees of alterations in cellular morphology, viability and adaptability ( Figure 7 and Table 2) (also seen in Figures B-E in Supporting Information). These study designs were aimed to select the optimal cytokine dosing and responsive microglial age group that could be utilized in in vitro conditions to develop a model culture system. Relied on the external morphology of the cells with maximal morphometric responsiveness as a primary indicator with a large combination of cytokines and their concentrations and age groups forming 80 options altogether in our study; the neonatal infants were favored for their maximal sensitivity for cytokine induction (Figure 7). Therefore, it was found that microglia from neonatal rats were very stable in primary culture and capable to respond and adapt much efficiently for longer in vitro than others in age continuum. Hence, they may be the potent candidate for studying different phenomena of the cell in variably conditioned ex situ environment. The less effectiveness in developing an ex situ platform for adult microglia, in contrary, had been an important finding which can be correlated with intrinsic cellular properties showing reduced proliferative potential and remarkable hike in ROS production leading to cellular stress in microglia with maturation (Figures 3 and 4).
This is directly correlated to the cytokine induced polarization state of the microglial population depending on IL-1β, IFN-γ, TNFα, IL-6 mediated 'classical activation' pathway which found more prevalent in aging with elevated oxidative stress from NADPH oxidase activated hydroxyl radicles, superoxides generation and NO production in various conditions (Czeh et al., 2011;Kim & Nagai, 2010;Pathipati & Ferriero, 2017).
Establishment of persistent microglial cell lines mostly from perinatal subjects, including a recent attempt of primary culture of adult microglia from a mouse actually taking sub-adult microglia along with other current studies on transcriptome data of the cell, jointly supports our notion of higher consistency and adaptability of microglia derived from very early life stage of rodent models with their suitability for primary culture as found in the present study (Galatro et al., 2017;Singh et al., 2014;Timmerman et al., 2018).

CONCLUSION
Therefore, microglial morpho-functional behavior throughout its ontogeny in the normal rodent brain showed us that their basic cellular properties have a specific pattern from perinatal stage to maturity. As we observed, in late embryonic and neonatal stages, microglia are not much dependent on ROS, which might be related with their functional responsibility of pruning and shaping CNS circuitry exerting minimal stress and now it has been supported with a current finding showing that they are using filopodial nibbling or 'trogocytosis' expressing PSD95 related to non-phagocytic membrane exchange (Weinhard et al., 2018). Intermediate young adult microglia settling in the CNS are tranquil in phagocytic behavior with low ROS production, but keen for surveillance and maintenance. Whereas microglia from matured adult rats tending toward aging brain showed high phagocytic vigor with dependence on cellular ROS and having lower proliferative potential. Microglial damaging behavior with age and in disease are the derivatives of their inherent behavioral changes in the normally maturing and aging brains with their polarization. Furthermore, the observed pattern of the basic cellular properties in the maturation continuum indicates the primary reasons behind the failure to extend adult cells ex situ for a longer period. In situ observations also provided explainable coherence with the documented patterns of cellular behavior of microglia through developmental and maturation continuum. As cytokine response behavior demonstrated better efficiency in neonatal infant microglia to change morphologically with cellular adaptability and behavior, they meet the requirements to perform as an effective ex situ model in rodents. By documenting changing microglial behavior through ontogeny, the present study has formulated a rapid isolation and culture of microglia, identifying its best age-matched responder group for working further on microglial biology. This study most importantly deciphered the shifting cellular properties of microglia with age, their differentiation toward the ROS dependent phagocytic mode from perinatal to matured stage only and found the best responding age for developing effective ex situ model for microglia.
Fellow partially. Her contribution has been specially acknowledged.
Authors acknowledge Ritesh Tiwari of BD-CoE Flow Cytometry unit at Centre for Research in Nanoscience and Nanotechnology (CRNN), University of Calcutta, for helping in acquiring and processing the data of FACS. Authors also expressing their gratitude to Jayasri Das Sarma, Indian institute of Science and Education Research -Kolkata (IISER-K) for extending her kind cooperation in organotypic brain slicing and related protocols. Payel Ghosh is thankful to University Grants Commission (UGC) for her research fellowship.

DATA AVAILABILITY STATEMENT
All data are represented in the manuscript and if required further, will be provided by the corresponding author.