Phagocytic Activity of Rat Primary Astrocytes Is Regulated by Insulin and Ganglioside GM1

A timely and efficient removal of apoptotic cells and their fragments is essential to maintain tissue homeostasis in normal and pathological conditions. Since the removal of apoptotic substrates is executed by the cells endowed with phagocytic activity, the issue of its regulation is of particular interest. In this work, we studied the effect of two biologically active substances, insulin and ganglioside GM1, on phagocytic activity of rat primary astrocytes. We showed that cell incubation with 1 µM insulin significantly decreased the phagocytic activity of astrocytes (58.5% vs. control), whereas the incubation with 10 µM GM1 caused an increase in phagocytic activity (133.4% vs. control). Preincubation of brain astrocytes with GM1 completely blocked the inhibitory effect of insulin. These results can be instrumental in developing novel therapeutic strategies for the treatment of neurodegenerative diseases accompanied by the emergence of apoptotic substrates.

Apoptosis is a vital process that occurs in all tis sues in normal and pathological conditions [1]. Apoptotic substrates, formed due to normal vital activity of cells in different organs, are promptly removed by specialized cells endowed with phagocytic activity. A timely and complete removal of apoptotic substrates is necessary to maintain tissue homeostasis, as well as to exclude the development of autoimmune reactions in response to intracellular antigens released from damaged and/or apoptotic cells [2].
Phagocytosis of apoptotic substrates is a com plex receptor mediated process in which the coordinated interaction of specific receptors on the plasma membrane of a phagocyte triggers sig naling mechanisms that ensure the ingestion and further cleavage of substrates by the intracellular lysosomal apparatus [3]. Thus, substrate recogni tion (binding), ingestion (engulfment), and cleav age are the main stages of phagocytosis that will be hereinafter referred to as stages 1, 2 and 3, respec tively. Moreover, the implementation of one of these stages of phagocytosis does not entail the automatic implementation of the next one.
Phagocytic function in the central nervous sys tem (CNS) is implemented by two cell types: microglial and astroglial cells. The former are cells of hematopoietic origin and refer to professional phagocytes, while the latter are non professional phagocytes [4]. Astrocytes are the most numerous neuroglial cells that account for up to 40% of the total population of brain cells [5]. These cells pro vide conditions for the generation and transmis sion of nerve impulses, thus contributing to the maintenance and implementation of synaptic function; they are involved in the formation of the blood-brain barrier and implementation of sup porting, trophic, secretory and phagocytic func tions. Astrocytes control not only the formation and function of synapses, but also their elimina tion in the developing and mature brain [6]. In pathological conditions of various etiologies (neu rodegenerative diseases, ischemia, aging, trau matic and radiation damage to the brain), the removal of apoptotic substrates and degenerative myelin by astrocytes takes on special significance [5].
As is currently known, the phagocytic activity of astrocytes is mediated by two receptor proteins of the plasma membrane: MEGF10 (Multiple EGF Like Domains 10) and the receptor tyrosine kinase MERTK (MER proto oncogene Tyrosine Kinase) [7]. The MEGF10 and MERTK proteins are involved in the removal of synapses, as well as damaged and/or apoptotic neurons formed during aging and/or under pathological condi tions of the CNS [5]. In addition, the MERTK receptor is assigned a key role in phagocytosis of myelin debris generated during natural processes of myelin renewal [8].
In tentative studies conducted in our labora tory, the process of phagocytosis of various apop totic substrates by rat brain primary astrocytes, as well as its kinetic characteristics, were thoroughly studied, and it was shown that phagocytosis of apoptotic substrates leads to an increase in the proliferative activity of brain astrocytes.
Astroglia proliferation is known to be an aggra vating factor in the development of pathological conditions of the CNS, accompanied by degenera tive changes and death of nerve cells. The accumu lation of astrocytes at the sites of local damage contributes to the development of astrogliosis, i.e. the process of replacing dead neurons with astro cytes. In this regard, the problem of optimization of astrocyte phagocytic activity (toward its decrease or increase) is fundamentally important for developing strategies to correct the pathological process, depending on the stage of its development.
In our laboratory, during the recent years, the neuroprotective effects of some natural biomole cules have been studied in vivo and in vitro, namely those of gangliosides and insulin during oxidative stress and a number of adverse effects on the organism [9]. Since the removal of apoptotic substrates is a vitally important factor that influ ences the survival of the CNS under unfavorable conditions, it appeared significant to find out whether these agents are able to regulate the phagocytic function of astrocytes.
Gangliosides, complex glycosphingolipids of cell membranes, are an integral component of the brain, and it has been shown in our laboratory that in mammals, their relative amount reaches 2.5 mg per 1 g of brain tissue (GM1) [10]. At the same time, it is well known that in degenerative and traumatic brain injuries, GM1 is released into the extracellular space, which suggests the possibility of its interaction with various brain cells, includ ing astrocytes. In addition, in our laboratory it was established that GM1 causes multiple activation of the phagocytic activity of professional phago cytes (macrophages) [11]. The brain is an insulin dependent organ, and various methods of insulin administration (intra cerebroventricular, intranasal, systemic) contrib ute to changes in eating behavior, as well as to weight loss, in experimental animals and humans [12]. Insulin receptors have been found on both neuronal and glial cells, including astrocytes [13,14]. Astrocytes are active components involved in insulin metabolism in the CNS. At the same time, astrocytes not only express the insulin receptor and insulin degrading enzyme [13,15], but are themselves insulin secreting brain cells [16], which indicates the importance of insulin signal ing for the normal functioning of these cells.
Nevertheless, in the available literature, we failed to find any data on the effect of insulin on the phagocytic activity of brain astrocytes. There is also no information on the possible modulating effect of gangliosides. Therefore, the goal of this study was to elucidate how insulin and GM1 affect the phago cytic activity of rat primary brain astrocytes.

Primary culture of astroglia
The studies were carried out on a primary cul ture of astroglia cells, which were isolated from the brain of newborn rats as described elsewhere [17]. The cells were cultured in a DMEM medium (Biolot, Russia) containing 10% inacti vated fetal calf serum (FCS; Biolot, Russia) and antibiotics (50 U/mL penicillin G and 50 μg/mL streptomycin; Biolot, Russia), on Petri dishes (Orange Scientific 3.5 cm), for 7−9 days at 37°C and 5% CO 2 until 75% confluence is reached. Cell viability was assessed by staining with Trypan Blue (Biolot, Russia), which is able to penetrate through damaged cell membranes.

Modeling of induced phagocytosis
As an apoptotic substrate of phagocytosis, we used the fraction of the rod outer segments (ROS) of the rat retina, which was isolated using a modi fied method described previously [18]. ROS preparations were conjugated to fluorescein iso thiocyanate (FITC; Merck, Germany) as reported in our previous publication [18]. The cells were washed with PBS (pH 7.4) and incu bated in a DMEM medium containing 2% inacti vated FCS. Phagocytosis substrates were then added at a cell/substrate ratio of 1:10. The number of cells and ROS fragments was counted using a Fuchs-Rosenthal chamber.
Cell treatment with modulators of phagocytic activity When analyzing the effect of insulin on the pro cess of phagocytosis, cells were incubated with 1 μm insulin (Sigma, USA) for 1 h, after which the FITC ROS conjugate was added. When studying the effect of GM1 on phagocytosis, cells were incubated for 24 h with the GM1 prepara tion at final concentrations of 0.1, 1 and 10 μm, after which the FITC NSP conjugate was added. GM1 was isolated from the bovine brain lipid extract obtained by the Folch method [19] as described previously [20].
When studying the combined effect of insulin and GM1, cells were incubated for 24 h with the GM1 preparation at final concentrations of 0.1, 1 and 10 μm, after which insulin was added at a concentration of 1 μm, and the cells were kept for 1 h. Next, the FITC ROS conjugate was added.

Analysis of phagocytic activity
After the end of the incubation with retinal ROS, the excess substrate was washed out with cooled PBS, and the cells were fixed for 3-5 min in 4% paraformaldehyde/PBS. After washing the prepa rations with PBS, the nuclei were stained with 0.001% Hoechst 33258/PBS (Serva, Germany). The experimental results were evaluated using a Zeiss Axio Imager A1 fluorescence microscope (Germany) equipped with an AxioCamMRc digi tal camera (Zeiss, Germany). ImageJ software (NIH, USA) was used to assess phagocytic activity; in doing this, we counted the number of cell nuclei (blue staining) and determined the relative area occupied by a phagocytosis substrate (green stain ing). The phagocytic activity of cells was assessed based on determining the implementation effec tiveness of phagocytosis stages 1 and 2. The quanti tative assessment of the process of substrate recognition and binding to astrocytes (stage 1) was carried out in arbitrary units by calculating the ratio of the fluorescent substrate area to the num ber of nuclei in the preparation. The morphologi cal criterion for assessing the substrate absorption by astrocytes (stage 2) was the presence of a fluo rescent FITC label in the perinuclear region, where the stained ROS fragments interact with lysosomes (the ratio of the number of cells contain ing a FITC labeled substrate in the perinuclear region to the total number of cells in the prepara tion was calculated). For quantitative analysis, cells were photographed, and two parallel slides were used for each point in the experiment. Fifteen randomly selected fields containing up to 1500 cells were photographed in each slide. Three inde pendent experiments were carried out. The results were statistically processed using a one way analy sis of variance (ANOVA), followed by Bonferroni's test for multiple comparisons, and GraphPad Prism software (San Diego, USA). Data are pre sented as M ± SEM.

Incubation with GM1 increases phagocytic activity
of primary rat brain astrocytes In the first series of experiments, we assessed the phagocytic activity of astrocytes in the pres ence of GM1 ganglioside. As a substrate of phago cytosis, we used rat ROS preparations, which represent a substrate of the apoptotic type [21]. In our previous studies, we have shown that ROS preparations are effectively recognized and ingested by rat brain astrocytes, with the kinetic parameters of ROS phagocytosis being compara ble to those characteristic of the capture and ingestion of apoptotic neurons. The results of the experiments are shown in Fig. 1 and Table 1. As follows from the data in Figs. 1a, 1a', in the con trol samples, the fluorescent signal was present both at the periphery and in the perinuclear region of cells, indicating the implementation of both phagocytosis stages (1 and 2). Figures 1b, 1b' and Table 2 show the results of experiments on assessing the phagocytic activity of rat brain astrocytes in the presence of GM1. We analyzed the effect of different GM1concentra tions ranging from 0.1 to 10 μm. It was found that due to incubation of astrocytes with GM1, the total amount of fluorescent substrate significantly increased compared to the control at GM1 con centrations ranging from 1 to 10 μm (Table 1). A quantitative assessment of the number of astrocytes containing a fluorescent substrate in the perinu clear area also revealed a significant increase com pared to the control (19.2 ± 2.7 and 19.5 ± 2.4 vs. 14.6 ± 1.5, or 131.5% and 133.6% of the control, respectively) (Fig. 1b, b'; Table 2). These data indi cate that GM1 stimulates the processes of both substrate binding (phagocytosis stage 1) and inges tion (phagocytosis stage 2) by astrocytes.

Incubation with insulin decreases phagocytic
activity of primary rat brain astrocytes After the incubation of astrocytes with insulin, the total amount of the FITC fluorescent label was statistically indistinguishable from the control (Figs. 1c, 1c'; Tables 1, 2). However, the amount of fluorescent label in the perinuclear region of cells preincubated with insulin was significantly reduced (to 58.5% of control), indicating a slow down in the implementation of phagocytosis stage 2 ( Fig. 1; Table 2). Indeed, the fluorescent substrate was predominantly localized at the periphery of astrocytes, associated with the plasma membrane. Thus, insulin turned out to be a potent inhibitor of the phagocytic activity of non professional brain phagocytes.
Since our laboratory is investigating a coaction of insulin and GM1 ganglioside, in the next series of experiments, we analyzed the joint effect of these natural biologically active molecules on the Effectiveness of phagocytosis stage 2 is assessed as a ratio of the relative number of cells containing ROS FITC conjugate in the perinuclear area to the total number of cells per preparation (%). (**)-Differences are significant at p ≤ 0.01. phagocytic activity of astrocytes. Figures 1d, 1d' show the results of a representa tive experiment, from which it follows that GM1 application led to a sharp increase in the fluores cent signal in the samples. Indeed, with GM1 concentrations of 1 and 10 μm, the amount not only of the bound substrate (Table 1), but also the absorbed substrate localized in the perinuclear region of astrocytes ( Fig. 1d; Table 2) significantly increased. The calculation of phagocytic activity showed that coincubation of astrocytes with GM1 blocks the inhibitory effect of insulin on the phagocytic activity of cells. At the same time, with a GM1 concentration of 1 μm, the inhibitory effect of insulin did not manifest itself, while the matching value of the amount of ingested sub strate reached the values characteristic of the con trol samples (15.4 ± 1.78 vs. 14.6 ± 1.5). Remarkably, with a GM1 concentration of 10 μm, the amount of substrate ingested by brain astrocytes (phagocytosis stage 2) was twice as high as the control values (26.6 ± 2.4 vs. 14.6 ± 1.5).

DISCUSSION
According to the recent data obtained in our laboratory on in vivo experimental ischemia in rats, the coadministration of insulin and GM1 ganglioside produced a unidirectional neuropro tective effect. These data are in good agreement with the results of the present in vitro study, in which the combined effect of these biologically active agents exceeded the stimulating effect of GM1 on the phagocytic activity of astrocytes.
According to the data obtained in the present work, GM1 applied at concentrations of 1 and 10 μm has a well pronounced ability to significantly increase the phagocytic activity of astrocytes when retinal ROS are used as an apoptotic substrate. At the same time, GM1 increased the effectiveness of both phagocytosis stages 1 and 2 (binding and ingestion, respectively). We are the first to demon strate here such effects of GM1 ganglioside. It is noteworthy that GM1 ganglioside is the most stable one of the 4 major brain gangliosides, which share similar neuroprotective effects and, as a rule, simi lar functional activity [22]. Therefore, it can be assumed that the main brain gangliosides may have the ability to activate the astroglia mediated pro cesses of phagocytosis, which is shown here using the effect of GM1 as an example.
The ability of gangliosides to activate phagocy tosis of apoptotic substrates by brain astrocytes, which we have established in this work, can be very valuable for developing novel strategies to treat various pathological conditions associated with the destruction of neurons and, probably, other brain cells. The destruction of cells in the central nervous system is characteristic of isch emic, traumatic and other brain lesions, including neurodegenerative diseases. It is well known that when brain neurons are destroyed, gangliosides, which are abundant in the brain, are released into the intercellular space [10]. Previously, we found that substances from destroyed brain cells first accumulate in the choroid plexus, and then are gradually excreted therefrom through the cerebro spinal fluid. Indeed, it has been demonstrated that in children with meningoencephalitis (postmortem studies), the level of gangliosides in the choroid plexus is higher than in the brain tissue [23]. Thus, if the level of gangliosides in the cerebrospinal fluid significantly increases during the destruction of neurons and other brain cells, then they could acti vate phagocytosis of apoptotic neurons by brain astrocytes, thus exerting a neuroprotective effect on the remaining viable brain neurons.
Although the mechanism behind the GM1 effect on the phagocytic activity of brain astro cytes is still unclear, the data on the effect of gan gliosides on receptor tyrosine kinases, which also include the phagocytic receptor MERTK, are of special interest. Furthermore, it is known that gangliosides can act through the tyrosine kinase Trk A receptor, thus activating the PI3K Akt1 mTOR signaling pathway [24,25]. As is known, mTOR activation leads to inhibition of autophagy [26]. In the case of non professional phagocytes, there is an inverse relationship between the pro cesses of autophagy and phagocytosis, i.e. activa tion of autophagy may entail a decrease in phagocytic activity. It is believed that this effect is due to the competition of both processes for the use of lysosomes, which finalize the cleavage of phagocytosis substrates and/or the process of autophagy in cells [27−29].
Besides, the in vitro beneficial effect of ganglio sides on the phagocytic activity of astrocytes can be enhanced by their coaction with insulin, which supports the data obtained in our laboratory in vivo. In our studies, insulin had an inhibitory effect on phagocytosis of apoptotic substrates by brain astrocytes. The molecular mechanism underlying this effect requires further research, while the data we have found in the available literature are scarce. For instance, similar results were obtained on the example of non professional phagocytes of bron choalveolar epithelial cells [30]. To explain this phenomenon, the authors pay special attention to AKT kinase in the regulation of phagocytic activity of non professional phagocytes of bronchoalveolar epithelial cells [30].
On the other hand, gangliosides are mainly localized in the membrane microdomains known as lipid rafts, being combined with other sphingo lipids and cholesterol [31]. The interaction of lipid rafts with membrane proteins plays an important role in cellular processes, such as signal transduction of cytokines, adhesion, intracellular transport, etc. [32]. Recently, it has been shown that on the plasma membrane of aortic endothe lial cells, which can also be considered as non professional phagocytes [33], GM1 ganglioside is co localized with the insulin receptor, while the disruption of the normal stoichiometric insulin receptor−GM1 relationship (towards an increase in GM1) leads to a blockade of insulin signaling [34]. It is noteworthy that in brain astrocytes, GM1 ganglioside is detected in insignificant amounts [35]; however, exogenously added GM1 can easily incorporate into the plasma membrane of astrocytes, leading to a modification of mem brane microdomains [36] and a change in the cel lular response to some stimuli [37]. In this context, our data on the abolition of the inhibitory effect of insulin in astrocytes that have undergone preliminary incubation with GM1 ganglioside are well consistent with the literature data.

CONCLUSION
In tissues delimited from the general circulation by histo hematic barriers, the removal of apoptotic substrates produced during their normal life is implemented by tissue resident cells exhibiting phagocytic activity. These cells promote the main tenance of tissue homeostasis and prevent the development of autoimmune reactions that may arise in response to the release of intracellular components from damaged or dying cells. In this work, we report for the first time data on two natu ral modulators exhibiting a cooperative stimulating effect on the phagocytic activity of non profes sional brain phagocytes, such as astroglial cells.

FUNDING
This study was carried out within a governmen tal assignment to the Sechenov Institute of Evolu tionary Physiology and Biochemistry (АААА А18 118012290427 7).

CONFLICT OF INTEREST
The authors declare that they have no conflict of interest, both evident and potential, that would be associated with the publication of this article.

OPEN ACCESS
This article is licensed under a Creative Com mons Attribution 4.0 International License, which permits use, sharing, adaptation, distribu tion and reproduction in any medium or format, as long as you give appropriate credit to the origi nal author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made. The images or other third party material in this article are included in the article's Creative Commons license, unless indi cated otherwise in a credit line to the material. If material is not included in the article's Creative Commons license and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this license, visit http://creativecommons.org/ licenses/by/4.0/.