Specialized Cortex Glial Cells Accumulate Lipid Droplets in Drosophila melanogaster

Abstract

Lipid droplets (LDs) are common organelles of the majority of eukaryotic cell types. Their biological significance has been extensively studied in mammalian liver cells and white adipose tissue. Although the central nervous system contains the highest relative amount and the largest number of different lipid species, neither the spatial nor the temporal distribution of LDs has been described. In this study, we used the brain of the fruitfly, Drosophila melanogaster, to investigate the neuroanatomy of LDs. We demonstrated that LDs are exclusively localised in glial cells but not in neurons in the larval nervous system. We showed that the brain's LD pool, rather than being constant, changes dynamically during development and reaches its highest value at the beginning of metamorphosis. LDs are particularly enriched in cortex glial cells located close to the brain surface. These specialized superficial cortex glial cells contain the highest amount of LDs among glial cell types and encapsulate neuroblasts and their daughter cells. Superficial cortex glial cells, combined with subperineurial glial cells, express the Drosophila fatty acid binding protein (Dfabp), as we have demonstrated through light- and electron microscopic immunocytochemistry. To the best of our best knowledge this is the first study that describes LD neuroanatomy in the Drosophila larval brain.

Fig 1. Lipid droplets in the Drosophila brain.

(A-E) Pictures taken from third instar larval brains. (A) Oil Red O staining, note an intense staining in the dorso-medial part of the central brain. (B) Toluidine blue-stained semithin section from the brain cortex. LDs (green) are organized in large clusters between neurons. (C) Pseudocolored electron micrograph taken from the brain cortex. LDs (green) are found in the cytoplasm of glial cells (G, red) but not in neurons (N). (D) High power electron micrograph from the perinuclear region of a cortex glia (G) LDs are rounded electron-opaque structures, delineated by a phospholipid monolayer (inset, arrow). N: neuron. (E) LDs (brown) are generally found in the closest vicinity of neuroblasts (asterisks). Unstained semithin section. (F) Diagram representing the distribution of LDs between neurons and glial cells. (G) Number of LDs in order of their distance from neuroblasts. (H) Time-course changes in the amount of accumulated LDs per brain tissue area ratio. Standard deviations are indicated. Scalebars: (A) 100 μm (B) 10 μm (C) 1μm, (D) 500nm, inset: 200nm, (E) 10 μm.

Fig 2. Cell type-specific distribution of LDs.

(A) Single cell clones of different glial cell types. Glial cell membranes are higlighted by myr-RFP (red) and LDs are highlighted by Lsd2-GFP (green). (B) Higher magnification image of a superficial cortex glia shown in panel A. (C) A superficial cortex glial clone (myr-RFP, red) on a pan-cortex glial cell labeled background (Nrv2-GFP, green). Note the similar morphology of the labeled glia as shown in panel B. (D) Pattern of LDs in the larval brain, visualized by ectopically expressed Lsd2-GFP shows similar distribution as the cortex glial specific GFP signal. (E) Box plot representing LD per cell area ratios from particular glial cell types. Scalebar: (A,B,C) 20 μm (D) 100 μm

Fig 3. Representative electron micrographs of particular glial cell types.

LDs are marked by arrows. (A) Perineurial (PG) and subperineurial cells (SPG) located on the surface of the brain. (B) A superficial glial cell located in the outer layer of the brain cortex, containing high amount of LDs. (C) A neuropil glia (NP) located at the cortex-neuropil boundary ensheating axons. (D) A deeper cortex glia (CG) found close to the cortex-neuropil boundary encapsulating a large peptiderg neuron (PN) and several other neurons (N) with its processes. No LDs seen in the cytoplasm of such a cortex glia. Note the presence of large clusters of neurosecretory vesicles (arrowheads) in the cytoplasm of PN. Scalebar: 2μm.

Fig 4. Ultrastructural features of the LD accumulating superficial cortical glial cells.

(A) HRP-labeled superficial cortical glial cell (CG₁), encapsulating a mitotic neuroblast (NB, arrows: chromosomes) and other neurons (N). Note the abundance of LDs. (A’) Same picture as shown in panel A but overexposed during acquisition to reveal the DAB stained processes (arrows) at low magnification. Note that a phagosome (ph, labeled in panel A) accumulates HRP. (B, B’) Higher magnification pseudocolored image taken from the dashed areas in panel A. DAB stained glial membranes are indicated (arrows). Two overlapping cortex glial processes (CG1 and CG2) insulate a mitotic neuroblast (NB). (C) Organization of the glial layers encapsulating neuroblasts. (C’) Subperineurial glial cells (SPG_1-2) are connected to each other through septate junctions. (C”) A subperineurial (SPG₂) and a superficial cortex glia (CG) are connected with an adherens junction. Scalebar: (A, A’, B) 1 μm, (B’,C, C’, C”) 100nm.

Fig 5. Light microscopic localisation of the Drosophila fatty acid binding protein (Dfabp) in third instar larvae.

(A) Western blot performed on total protein extracts from dfabp ³²⁵² homozygous mutant and wild type larvae. The Dfabp antibody labels a single band at 14kDa in the wild type sample, while no labeling is observed in samples from mutants. Immunohistochemistry on wild type (B), and on dfabp RNAi (C) third instar larval brains using the Dfabp antibody (green). Nuclei are stained with DAPI (blue). Note the absence of staining in RNAi silenced compared to wild type animals. (D) Higher magnification image of the dorsomedial part of the central brain. The Dfabp antibody reveals a thin network, between neuroblasts (asterisks) and their daughter cells. Two Dfabp-positive (arrowheads), and one unlabeled soma (arrow) are visible at the brain surface. (E) Double labeling against Dfabp (green) and the glial-specific protein Repo (red). Note that Dfabp is present in the cytosol and in the nucleus of glial cells (arrows). (F) Double labeling for cortex glia-GFP (green) and Dfabp (red). Perineurial cells (arrow) are double negative for Dfabp and GFP. Subperineurial cells (arrowhead) are positive for Dfabp and negative for GFP. Cortex glial cells (double arrow) are double positive for Dfabp and GFP. Scalebar: B, C:100 μm; D, E, F, F’: 10 μm.

Fig 6. Ultrastructural localisation of Dfabp.

Post-embedding silver-intensified immunogold labeling for Dfabp on freeze-substituted LR White-embedded material. An intense labeling can be seen over subperineurial (SPG) and superficial cortex (CG) glial cells. Note the absence of labeling over perineurial glia (PG), neurons (N), or neuroblasts (NB). Arrows: lipid droplets, arrowheads: mitochondria. Scalebar: 1 μm

Fig 7. Schematic illustration of the relative distribution of lipid droplets between glial cell types and the localization of Dfabp.

LDs are concentrated in large clusters in the perinuclear region of glial cells but are not present at all in neurons (N). Specialized superficial cortex glial cells (CG) insulating neuroblasts (NB) and their daughter cells (asterisks) accumulating the highest amount of LDs. Neighboring subperineurial cells (SPG) establish septate junctions (SJ), while SPG and superficial cortex glial cells are connected to each other through adherens junctions (AJ). The Drosophila fatty acid binding protein (Dfabp) is expressed in LD accumulating superficial cortex glial cells and subperineurial (SPG) cells, and is localized in the cytosol and in the nucleus. NL: neural lamella, PG: perineural glia, NG: neuropil glia, ax: axon.