INTRODUCTION

Glia is a system of cellular elements in the nervous system that ensures structural stability and functioning of nerve cells. Similarly to other animals, the glial system of insects includes cells that differ morphologically, topographically, and functionally. In the modern literature, insect glial cells are divided into four main types: perineurium, subperineural glia, cortical glia, and neuropilar glia; the latter is subdivided into ensheating and astrocyte-like glia. In addition, a class of glial cells that accompany nerve fibers in nerve roots is distinguished (Edwards and Meinertzhagen, 2010; Freeman, 2015). Sometimes, in the classification of glial cells, more minor characteristics are taken into account, and the number of distinguished types of glial cells may increase several times (e.g., up to 14 in the first thoracic ganglion of the tobacco hornworm Manduca sexta (Cantera, 1993)).

The cortical glia, which will be discussed in this article, is an assemblage of glial cells confined to the layer of neuronal bodies in the brain and trunk ganglia. It presumably performs two main functions: the formation of a supporting framework, which encloses the bodies of neurons, and the implementation of a number of metabolic functions (Freeman, 2015). This is especially evident in the case of large motoneuron bodies, which form “trophospongy,” a system of invaginations of the plasma membrane of neuron bodies, which include thin processes of cortical glial cells (Wigglesworth, 1959).

Molecular genetic methods for the selective detection of cortical glial cells were developed only for Drosophila (Awasaki et al., 2008). For other insects, these methods are inapplicable. Using conventional histological staining methods, only cell nuclei and part of the perinuclear cytoplasm can be identified clearly. For this reason, the topographic and morphological characteristics of cortical glial cells are given primarily on the basis of their nuclei.

At least two dimensional types of cortical glial cells are known. Cells with small round or oval nuclei several micrometers in diameter were found, for example, in Periplaneta americana (Scharrer, 1939) and Drosophila melanogaster (Hartenstein et al., 2008). On the other hand, in a number of insects, “giant” cells with large polymorphic nuclei were found, which, judging by their position, can be attributed to cortical glia. Among the hemimetabolous insects, “giant” cells were found in the bug Rhodnius prolixus (Wigglesworth, 1959); among the representatives of Holometabola, these cells were found in the honey bee (Weyer, 1931; Risler, 1954), two species of beetles (De Lerma, 1949), the housefly (Grandori et al., 1951), and several lepidopteran species (Schrader, 1938; Panov, 1963; Nordlander and Edwards, 1969; Cantera, 1993). For some species, the sizes of giant cell nuclei were specified: up to 90 µm in Hydrous piceus (De Lerma, 1949), 60 × 15 µm in the Rhodnius bug, but only 35 µm in Oncopeltus fasciatus and less than 30 µm in the house fly (Grandori et al., 1951).

However, these studies do not allow a sufficiently objective assessment of the extent of distribution of this phenomenon in Insecta. Therefore, the goal of this work was to search for giant cortical cells in a possibly wider range of insects and to identify their features in different representatives of this class. Taking into account the data of previous studies, we determined the conditional size boundary between the “small” and “giant” nuclei of cortical glial cells at the level of ~30 µm. The examination was conducted in the supraoesophageal and suboesophageal ganglia, except for the optic lobes, in which the system of glial cells is arranged in a special way and is considered separately from that of the “central” brain (Nordlander and Edwards, 1969; Tix et al., 1997).

MATERIALS AND METHODS

Series of sections of the brain of 152 species from 13 orders of hemi- and holometabolous insects from the collection of preparations of the author were studied (Table 1). The insects were collected during long-term expeditionary and laboratory work by the author and other entomologists of the Severtsov Institute of Ecology and Evolution, Russian Academy of Sciences. The incised insects were fixed in Bouin’s fluid, the dissected supraesophageal ganglia were embedded in the paraplast, and a series of sections 7–10 µm thick were stained with Heidenhain’s iron haematoxylin (Romeis, 1953). On the resulting sections, the nuclei of glial cells are clearly seen. On the basis of their position, various types of glial cells were identified.

Table 1. List of species studied
Full size table

RESULTS

In the mayflies, cockroaches, orthopterans, phasmopterans, earwigs, and cicadas studied (Table 1), the cortical glia in the layer of brain neuron bodies was represented only by small angular cells with rounded or ellipsoidal nuclei containing chromatin in the form of intensely stained granules. Their diameter was comparable to or even smaller than the diameter of the nuclei of neighboring neurons (Fig. 1a, Table 2). After staining using the procedure described, the cytoplasm either was not detected at all or was seen in the form of thin dark-colored strands extending from the perinuclear region and going between the bodies of neurons. Thus, the cortical glial cells of insect species of the specified groups were similar to those described previously in the American cockroach (Scharrer, 1939).

Fig. 1.
figure 1

Cortical glial cells in the brain of insects: (a) Acheta domesticus, (b) Rhodnius prolixus, (c) Triatoma maculate, (d) Triatoma infestans, (e) Coreus marginatus, (f) Chrysopa perla. Designations (for Figs. 1–3): CG, cortical glial cell nuclei; GCG, “giant” cortical glial cell nuclei; GCGc, “giant” cortical glial cell cytoplasm; KC, Kenyon cells of mushroom bodies; TR, tracheoblast nucleus. Scale bar (Figs. 1–3), 10 µm.

Full size image
Table 2. Maximum cross-sectional size of cortical glial nuclei (insect species with “giant” cortical glial nuclei are shown in bold)
Full size table

Since the giant cortical glia was found in the hemipteran R. prolixus, representatives of the order Heteroptera were studied in more detail. In total, sections of the supraesophageal ganglion of 17 species from ten families were at our disposal (Table 1). As a result of analysis, giant glial nuclei were found in the layer of bodies of brain neurons in only three species of the Reduviidae family. First, they were found in R. prolixus, which confirms their existence in the central nervous system (CNS) of this bug (Wigglesworth, 1959). However, their content in the brain is, apparently, lower than in the trunk synganglion studied by Wigglesworth. In particular, they were found only in the dorsal protocerebrum, where their nucleus, filled with granular chromatin, surrounded in a cap-like manner the assemblage of Kenyon cells of the mushroom body (Fig. 1b). Second, the characteristic giant nuclei were also found in the brain of Triatoma infestans and Triatoma maculata. In addition to the nuclei that separated Kenyon cells from neighboring neurons, other giant nuclei were found in the cell layer of the brain of both Triatoma species. The distribution and structure of chromatin in the giant nuclei in these species differed. In Rodnius, chromatin grains were smaller (Fig. 1b), whereas in the species Triatoma, chromatin aggregated into clumps, which were smaller in T. maculata (Fig. 1c) and larger in T. infestans (Fig. 1d).

In hemipterans of 14 other species studied, representing eight families, no giant nuclei were found in the cortical layer of the brain. In only three species (Dicranocephalus agilis from the family Stenocephalidae and Enoplops scapha and Coreus marginatus from the family Coreidae), the size of some nuclei of cortical glial cells was slightly increased but did not exceed the conditional boundary separating the “small” and “giant” cells of the cortical glia (Fig. 1e).

In the highly taxonomically diverse order Neuroptera, only the brain of adult Chrysopa perla was examined. In the cell cortex, among the bodies of neurons, only angular nuclei with granular chromatin, comparable in size to the nuclei of neighboring neurons, were found (Fig. 1f).

In beetles, the sizes of the nuclei of cortical glial cells varied significantly. For example, the presence of “giant” nuclei of cortical glia, which were first described by De Lerma (1949), was confirmed in Hydrophilus piceus. However, in Hydrobius fuscipes and Sphaeridium scarabaeoides from the same family Hydrophilidae, as well as in Nicrophorus vespilloides from the family Silphidae, the cortical glial cells were small. The cortical glial cells in the studied representatives of ground beetles, dark beetles, and barbels were also small.

To reveal a possible correlation between the size of cortical glial cells and the taxonomic position of the species, we studied supraoesophageal ganglia of 34 beetle species from five families of the superfamily Scarabaeoidea (Table 1). In general, with some exceptions, the “lower” scaraboids have small cortical glial cells (Fig. 2a). Exceptions to this trend are representatives of Passalidae (Table 2) and Glaphyridae. In an unidentified representative of Passalidae from southern Vietnam, among the bodies of neurons in the cell cortex, large nuclei of glial cells occurred, the sections of which reached ~40 μm in length (Fig. 2b). In Amphicoma vulpes, the sections of the nuclei of cortical glial cells differed significantly in size: among the bodies of neurons, both small nuclei and nuclei with a diameter equal to approximately three diameters of neighboring neurons could be seen.

Fig. 2.
figure 2

Cortical glial cells in the brain of insects (continued): (a) Geotrupes stercorarius, (b) unidentified beetle from the family Passalidae, (c) Amphimallon volgensis, (d) Polyphylla fullo, (e) Apis mellifera, (f) Lymantria dispar.

Full size image

Among the representatives of the family Scarabaeidae, the studied species of the subfamilies Aphodiinae and Scarabaeinae contained small cortical glial cells. In addition, cortical cells were also small in many herbivorous scarabids (Serica brunnea, Miltotrogus aequinoctialis, Amphimallon volgensis, Hoplia parvula, Adoretus nigrifrons, Pentodon algerinum bispinifrons, Trichius fasciatus, Oxythyrea funesta, and Tropinota hirta) (Fig. 2c). Medium-sized cortical cells were found in a smaller part of representatives of Scarabaeidae (Melolontha hippocastani, Polyphylla fullo, Anoxia pilosa, Phylopertha horticola, Anisoplia agricola, Pentodon algerinum bispinifrons, Cetonia aurata, and Protaetia metallica). The largest nuclei were found in P. fullo and A. pilosa (Fig. 2d, Table 2). However, the “true” giant cells, which will be further described in Lepidoptera, were found in none of the cases.

The hymenopterans studied contained primarily small cortical glial cells. The only exception is the honey bee, in which larger glial nuclei were found in the cell cortex of the brains of some specimens (Fig. 2e).

In the only representative of scorpionflies examined, Panorpa communis, the cortical glia was small-celled.

In representatives of the order Lepidoptera, as noted above, the presence of giant polymorphic nuclei in the layer of bodies of brain neurons and trunk ganglia was demonstrated most convincingly. To test the assumptions about the gigantism of cortical glial cells as a feature characteristic of at least higher Lepidoptera in general, in this work we studied the brain sections of 15 lepidopteran species from nine families (Table 1). In all specimens, giant nuclei of cortical glial cells characteristic of Lepidoptera were found. As in the species studied previously, they were located partially under the layer of subperineural glia and partially in the depth of the cell cortex. Chromatin had a granular structure and filled the nucleus evenly or aggregated in the center. Large nucleoli were present; however, they were, apparently, scanty, because they were not found in all nuclear sections (Fig. 2f).

The closest relatives of lepidopterans are caddisflies, which together with them form the superorder Amphiesmenoptera (Kiriakoff, 1948). Therefore, in connection with the wide distribution of gigantism of cortical glial cells in Lepidoptera, it was of interest to determine the characteristics of cortical glia in caddisflies as well. We had three species of caddisflies at our disposal (Table 1). In all of them, giant nuclei were present in the layer of neuron bodies. They, as it were, delimited the cell cortex from the brain sheath or penetrated between the bodies of neurons, not differing in shape and structure from the giant nuclei of the cortical glial cells of Lepidoptera (Fig. 3a, Table 2).

Fig. 3.
figure 3

Cortical glial cells in the brain of insects (continued): (a) Limnephilus flavicornis, (b) Hydrotaea ignava, (c) Muscina stabulans, (d, e) Muscina levida, (f) Limonia quadrimaculata.

Full size image

For Diptera, two dimensional types of cortical glial cells are currently known: the Drosophila type, which is characterized by a small cell size, and the Musca type, with large (“giant”) glial cells (Grandori et al., 1951). In the housefly brain, we also found large rounded nuclei with chromatin clumps, often adjacent to the nuclear envelope, and a large nucleolus. These nuclei were located in the depths of the cell cortex and near the perineurium. They were surrounded either by a layer of fibrous cytoplasm or by the bodies of neighboring neurons closely adjoining them.

In addition to the housefly, similar nuclei of cortical glial cells were also found only in other members of the family Muscidae (Table 1) (Fig. 3b), as well as in Fannia scalaris (Fanniidae) and Scatophaga stercoraria (Scatophagidae), i.e., in representatives of families belonging to the superfamily Muscoidea.

The structure and location of the giant nuclei of cortical glia have been studied in most detail in three species of Muscina. As in the housefly, nuclei were found both in the periphery of the cell cortex layer and inside it. In the peripheral nuclei, the cytoplasm surrounding them was clearly seen (Fig. 3c). Previously, the spatial relationship of giant nuclei with secondary neuroblasts of mushroom bodies was noted (Panov, 2020).

In the 3rd instar larva, the nuclei of the cortical glial cells had very large nucleoli (Fig. 3e), and the average diameter of the nuclei was ~20 µm. In prepupae and pupae, the size of the nuclei increased, whereas in adults it could reach nearly 30 µm. At all developmental stages, the nuclei were rounded and contained one nucleolus, the size of which was maximum in larvae and decreased in pupae and adults. On the equatorial sections of the nuclei, the chromatin looked like intensely stained, clearly defined large clumps located adjacent to the nuclear envelope (Fig. 3c). On the tangential sections of the nuclei, passing through their “humps,” it could be seen that these chromatin clumps are fragments of bead-shaped structures, polytene chromosomes (Fig. 3d). In all three species of Muscina, no divisions of the nuclei of cortical glial cells were found either in larvae or during metamorphosis.

In all the remaining 40 species belonging to 27 families of long- and short-horned dipterans (Table 1), the cortical glia was small-celled (Fig. 3f) and giant nuclei of the cortical glia were not found.

DISCUSSION

The study of the brain of 152 species, representing 13 most numerous orders of the class Insecta in terms of species composition, showed that the primary size type of cortical glia is the small-celled glia, the morphological, functional, and molecular genetic characteristics of which have been studied in recent years in Drosophila (Hartenstein, 2011). In the trunk synganglium of Drosophila, cortical glial cells (6–8 per neuromer) are formed as a result of the division of four neuroglioblasts in embryonic time (Altenhain, 2015). The same number of cortical glial cells is retained in late embryos and newborn larvae; later, the cells begin to divide mitotically, and their number increases significantly. In the house cricket Acheta domesticus, which also has small-celled cortical glia, its cells in nymphal time divide mitotically, similarly to the cells of neuropilar glia, in the rhythm of molts. The peak of mitotic activity is observed at the end of the third and the beginning of the fourth quarter of the intermolt period (Panov, 1962), when the titer of the molting hormone in insects reaches its peak.

This study of a wide range of insects showed that the “gigantism” of the cortical glial cells is a phenomenon that is observed rather rarely either in individual species or in taxonomic groups of different ranks. Two types of gigantism can be distinguished. The first type, which is characterized by very large complex-shaped nuclei filled with numerous small chromatin grains, is most fully represented in Lepidoptera. Such giant nuclei were first described in the mill moth (Schrader, 1938); however, the author considered them to be the nuclei of tracheolar cells. Later, in the Chinese tussar moth Antheraea pernyi, giant nuclei of two categories differing in structure were found in the cell layer of the trunk ganglia and brain. Some of them, indeed, were the nuclei of tracheolar cells, whereas others were the nuclei of giant cortical glial cells (Panov, 1963).

In the Chinese tussar moth embryos, at the time of the formation of the neurilemma, sporadic elliptical nuclei were found immediately below the meurilemma and among the bodies of neurons. At later stages of embryogenesis, these nuclei began to differ from the nuclei of nerve cells in irregular contours and a uniform distribution of chromatin grains in the nucleus. Divisions of these nuclei were observed neither in embryos nor in caterpillars. However, they grew intensively and became the largest cell nuclei in adult caterpillars both in the brain and in the trunk ganglia.

Only one period of division of giant glial cells (in the first days of pupal development) was established. At that time, figures of huge mitoses at the metaphase stage were found in a fairly large number of cells in the surface layer of the ganglia (Panov, 1963).

Later, the existence of giant cortical glial cells in Lepidoptera was confirmed when studying the postembryonic development of the nervous system in the monarch butterfly Danaus plexippus plexippus. In the latter, giant cortical glial cells grow without division throughout the larval period of development but enter mitosis at the beginning of metamorphosis (Nordlander and Edwards, 1969). Similarly to the Chinese tussar moth, giant mitoses in the monarch butterfly were found only in the metaphase. For this reason, the further fate of the giant cortical glial cells that entered mitosis remained unknown.

Giant cortical glial cells were also found in the prothoracic ganglion of the tobacco hawk moth, which entered mitosis during the period of metamorphosis, and even multipolar mitoses were observed. A simultaneous increase in the number of giant nuclei, which was accompanied by a decrease in their size, suggested that karyokinesis ends with the formation of daughter nuclei (Cantera, 1993).

There is ample literature on the problem of the correlation between the size of nuclei and the degree of their ploidy, which contains evidence of both the presence and absence of such a correlation (Brodsky and Uryvaeva, 1981). Modern reviews state the existence of this correlation (Edgar and Orr-Weaver, 2001). The large size of the nuclei of giant cortical glial cells and the abundance of chromatin material in them suggested their polyploid nature (Johansson, 1957; Panov, 1963; Nordlander and Edwards, 1969). This assumption was confirmed by Nordländer and Edwards, who observed multiple cycles of H3-thymidine incorporation into the giant nuclei of the cortical glial cells in the monarch butterfly during the larval period without their subsequent divisions.

Cases of instrumentally corroborated polyploidy of insect CNS glial cells are rare. For example, it was recently discovered that the cells of the subperineural glia of D. melanogaster, which is part of the blood–brain barrier, are polyploid, and an increase in the size of subperineural glial cells as a result of polyploidization is required to maintain the integrity of the intercellular junctions that form the blood–brain barrier (Unhavaithaya and Orr-Weaver, 2012). In the CNS of adult Drosophila, an increase in the number of cells with a DNA content from 4C to more than 16C with aging of the insect was found. Identification of the cell type in preparations of the dissociated central nervous system using molecular markers showed that, for two weeks of imaginal life, 6–7% glial cells in general begin to contain more than 2C DNA (Nandakumar et al., 2020).

Divisions of polyploid cells were found in the gut of dipteran larvae. Fox et al. (2010) detected mitotic divisions of octoploid cells in the hindgut of a metamorphosing Drosophila larva. On the basis of this, they concluded that the entry of cells into the endocycle is not irreversible and does not rule out the possibility of their subsequent reproduction by mitosis. Apparently, this is also true for the presumably highly polyploid cortical glial cells of Lepidoptera described in this study.

A second, specific type of polyploidy is polyteny. In this case, the chromonemas arising during repeated endocycles do not diverge and remain closely connected, which leads to the formation of giant polytene chromosomes (Stormo and Fox, 2017). Cells of a number of internal organs of some species of dipterans and other insects are polytene (Brodsky and Uryvaeva, 1981).

On histological sections, polytene chromosomes usually look like large intensely stained clumps of chromatin that tightly adjoin the nuclear envelope. Only on the surface sections of the nuclei can it be seen that these clumps are combined into bead-like chains—polytene chromosomes. Until now, there has been only one bit of evidence of the presence of polytene cells in the nervous system of insects. In Drosophila virilis, polytene nuclei were found in squashed preparations of the trunk synganglion of the larva, but their affiliation to any particular cell type was not established (Makino, 1938). In fact, polytene chromosomes in housefly glial cells were first described by Grandori et al. (1951); however, they were not qualified as such. Their presence in the nuclei of cortical glial cells is highly likely in several species of the muscoid Diptera (the third case of detection of polytene cells in the central nervous system of insects).

In insects, the majority of organs containing cells with polytene chromosomes are “larval,” degenerating during metamorphosis. The only exceptions are Malpighian vessels and ovarian trophocytes (Brodsky and Uryvaeva, 1981). These metamorphosing structures with polytene chromosomes apparently also include the cortical glial cells described in the muscoid Diptera in this work. However, there was no one case of finding mitoses in them during postembryonic development. The impossibility or great rarity of the entry of cells with polytene chromosomes into the mitotic cycle is explained by the peculiarities of the structure of polytene chromosomes and errors during genome reduplication in cells of this type (Edgar and Orr-Weaver, 2001).

This work only outlines the general picture of the distribution of cell gigantism in the cortical glia of insects and, most likely, the underlying polyploidy. It remains unknown why, in certain groups of insects of different taxonomic ranks (from subfamily to superorder), a transition from the proliferation of initial cellular elements to a significant growth of cortical glial cells without their multiplication is observed. It is generally believed that polyploidization, which leads to cell growth and increase in their synthetic potential, saves time and resources in comparison with growth through proliferation (Brodsky and Uryvaeva, 1981). Compared to mitotic cycles, which are accompanied by destruction of the nuclear envelope, spindle formation, cytokinesis, nuclear envelope repair, etc., endoreplication preserves the structure and specific functioning of polyploidizing cells. For example, the polyploidization of subperineural glial cells in Drosophila, which forms the blood–brain barrier, is explained by the need for continuous specific functioning (Unhavaithava and Orr-Weaver, 2012).

In the case of the cortical glia, the functional and morphogenetic causes for the cases of its gigantism and the presumably polyploid nature remain unknown. It is interesting to note that the neuropilar glia is always small-nucleated. It remains unclear whether the gigantism of the cortical glia is related to the general trend of morphogenetic processes in the nervous system of insects of various groups. Only single finds suggest the possibility of such a relationship. For example, when studying the composition of brain neurosecretory cells, two groups of insects were distinguished. In insects of the first group, the number of neurosecretory cells increased during postembryonic development due to the activity of neuroblasts. In insects of the second group, the medial neurosecretory cells of the brain were relatively scanty, and their number was mainly stabilized as early as at the beginning of postembryogenesis. It is interesting to note that the neurosecretory cells in the studied insects of the first group remained diploid, whereas in one of the adult insects of the second group they were tetraploid (Panov and Marshak, 1968). The insects with giant cortical glial cells are representatives of insects with a stable composition of medial neurosecretory cells of the brain.