Non-Autonomously Regulated Autophagy
-
1.
Non-autonomously regulated autophagy in Aging
Aging is best described as a progressive increase in the probability of death for an individual with the passage of time [
72,
73]. It is well established that there is a gradual decline of autophagy leading to loss of proteostasis with age, and hence, loss of autophagy is often considered one of the causal factors of aging; decreasing proteostasis is systemically controlled among tissues, hence suggesting a cell non-autonomous regulation of autophagy [
74,
75,
76,
77,
78,
79,
80].
Studies in
D. melanogaster have established that signaling pathways like mTOR, EGFR, and AMPK regulate autophagy [
2,
3,
81]. AMPK signaling in fruitflies has anti-aging effects when neuronally overexpressed, resulting in lifespan extension. Overexpression of AMPK in neurons using Elav-GS led to autophagy induction locally (within the tissue), but amazingly, it also led to a significant autophagy induction non-autonomously regulated in the intestinal cells and muscle cells. Induction of autophagy within the intestine prevented aging-associated intestinal dysfunction. Likewise, autophagy induction in muscles led to improved muscle function during aging. In yet another interesting experiment, AMPK overexpression within the intestine led to autophagy induction in the brain non-autonomously regulated, suggesting that there is crosstalk between these tissues to regulate autophagy [
4,
82]. Thus, the gut-brain axis is involved in the cross-regulation of autophagy in these tissues. The data further show that slowed intestinal aging led to increased organismal lifespan, and aging in muscle cells was also delayed due to the sequestration and subsequent destruction of damaged proteins and organelles. The researchers demonstrated that the beneficial effects of AMPK/Atg1 overexpression are due to a systemic increase in 4E-BP expression and reduced DILP (
Drosophila Insulin-like peptides) levels in the brain [
4]. In addition, overexpression of the uppermost hierarchical autophagy kinase Atg1 in the brain distally induces autophagy in midgut enterocytes, which assists in maintaining intestinal homeostasis [
4]. The model for this regulation is shown in
Figure 3A.
Figure 1.
Autophagy and its regulation. Within the cytoplasmic milieu of eukaryotic cells, induction of autophagy leads to the formation of double-membrane organelles, known as autophagosomes. Autophagosomes carry cytoplasmic cargo of toxic proteins, dysfunctional organelles, and intracellular pathogens such as bacteria and viruses. This orchestrated process, termed autophagy, is initiated by the ULK1 complex, a multimeric assembly comprising ULK/ Atg1, ATG13, FIP200, and ATG101. Vps34 complex consists of Vps34, Vps15, Atg6 and Vps14. WIPI-like proteins, including Atg18, bind phosphoinositides to recruit Atg2 and Atg9. Autophagosome elongation and maturation entail the concerted action of two ubiquitin-like conjugation systems: the LC3/Atg8 (Atg8, Atg4, Atg7, Atg3) and ATG12 (Atg12, Atg5, Atg16, Atg7, Atg10) pathways. Upon closure, autophagosomes undergo irreversible fusion with lysosomes (catalyzed by SNAREs, HOPS complex, Atg14 etc.) culminating in the formation of autolysosomes. Within these acidic compartments, macromolecules undergo enzymatic degradation, yielding amino acids, fatty acids, monosaccharides, and nucleotides, which are subsequently recycled into the cytosol, thus sustaining cellular metabolism and homeostasis.
Figure 1.
Autophagy and its regulation. Within the cytoplasmic milieu of eukaryotic cells, induction of autophagy leads to the formation of double-membrane organelles, known as autophagosomes. Autophagosomes carry cytoplasmic cargo of toxic proteins, dysfunctional organelles, and intracellular pathogens such as bacteria and viruses. This orchestrated process, termed autophagy, is initiated by the ULK1 complex, a multimeric assembly comprising ULK/ Atg1, ATG13, FIP200, and ATG101. Vps34 complex consists of Vps34, Vps15, Atg6 and Vps14. WIPI-like proteins, including Atg18, bind phosphoinositides to recruit Atg2 and Atg9. Autophagosome elongation and maturation entail the concerted action of two ubiquitin-like conjugation systems: the LC3/Atg8 (Atg8, Atg4, Atg7, Atg3) and ATG12 (Atg12, Atg5, Atg16, Atg7, Atg10) pathways. Upon closure, autophagosomes undergo irreversible fusion with lysosomes (catalyzed by SNAREs, HOPS complex, Atg14 etc.) culminating in the formation of autolysosomes. Within these acidic compartments, macromolecules undergo enzymatic degradation, yielding amino acids, fatty acids, monosaccharides, and nucleotides, which are subsequently recycled into the cytosol, thus sustaining cellular metabolism and homeostasis.
Figure 2.
Cell-autonomous and cell non-autonomous regulation of autophagy. Cell-autonomous autophagy occurs within the cytoplasm of the cell upon several stimuli which can culminate on one or more of the proteins shown. AMPK, FOXO and exercise positively regulate autophagy while nutrients, insulin, AKT and mTOR negatively regulates autophagy. Cell non-autonomous autophagy refers to a phenomenon in which the autophagic activity of a cell is influenced by signals or factors originating from neighbouring cells or tissues. This could occur by secreted factors and membrane bound proteins. The induction of cell non-autonomous autophagy can occur in neighbouring cell (green) or a distant cell/tissue (light blue).
Figure 2.
Cell-autonomous and cell non-autonomous regulation of autophagy. Cell-autonomous autophagy occurs within the cytoplasm of the cell upon several stimuli which can culminate on one or more of the proteins shown. AMPK, FOXO and exercise positively regulate autophagy while nutrients, insulin, AKT and mTOR negatively regulates autophagy. Cell non-autonomous autophagy refers to a phenomenon in which the autophagic activity of a cell is influenced by signals or factors originating from neighbouring cells or tissues. This could occur by secreted factors and membrane bound proteins. The induction of cell non-autonomous autophagy can occur in neighbouring cell (green) or a distant cell/tissue (light blue).
Figure 3.
Cell non-autonomous autophagy in aging. A. Increased levels of AMPK and Atg1 in D. melanogaster adult brain (grey) led to local upregulation of autophagy and reduced DILPs, thus delaying aging locally. Interestingly, this also results in enhanced autophagy cell non-autonomously in intestine and muscles (intestinal cell: yellow, muscle cell: blue) resulting in improved intestinal homeostasis and improved muscle function thus, delaying systemic aging. Further, AMPK overexpression in intestine resulted in autophagy upregulation in brain and muscles, thereby, delaying systemic aging and improved muscle function during aging. B. Aging results in accumulation of miR-83 which inhibits autophagy through CUP-5 and thus, progressing aging. When C. elegans is mutated for miR-83, autophagy is upregulated in BWM (green), which do not express miR-83, suggesting transport of miR-83 from intestine (yellow) to the BWM, thus regulating autophagy cell non-autonomously.
Figure 3.
Cell non-autonomous autophagy in aging. A. Increased levels of AMPK and Atg1 in D. melanogaster adult brain (grey) led to local upregulation of autophagy and reduced DILPs, thus delaying aging locally. Interestingly, this also results in enhanced autophagy cell non-autonomously in intestine and muscles (intestinal cell: yellow, muscle cell: blue) resulting in improved intestinal homeostasis and improved muscle function thus, delaying systemic aging. Further, AMPK overexpression in intestine resulted in autophagy upregulation in brain and muscles, thereby, delaying systemic aging and improved muscle function during aging. B. Aging results in accumulation of miR-83 which inhibits autophagy through CUP-5 and thus, progressing aging. When C. elegans is mutated for miR-83, autophagy is upregulated in BWM (green), which do not express miR-83, suggesting transport of miR-83 from intestine (yellow) to the BWM, thus regulating autophagy cell non-autonomously.
MicroRNAs (miRNAs) are identified as an important regulator of many biological phenomena, including aging and autophagy [
83,
84]. Interestingly, a few miRNAs have been shown to be secreted out of the source cell into the body fluids [
85]. For instance,
mir-83 in
Caenorhabditis elegans, a homolog of mammalian
mir-29, is differentially expressed in neurons and the intestines, and its expression is regulated in an age-dependent manner through heat shock factor-1, the master transcription factor of the heat shock response [
86,
87].
mir-83 impairs autophagy cell-autonomously by supressing cup-5 in the intestinal cells and it modulates autophagy in the body wall muscles (BWM) non-autonomously regulated also via Cup-5 (
Figure 3B).
mir-83-/- mutants showed increased autophagosomes and autolysosomes in the BWM, even though BWM does not express
mir-83.
mir-83 regulates autophagy in BWM non-autonomously through cup-5, as disruption of cup-5 in BWM via RNAi restored autophagy to normal levels [
87]. Further,
mir-83 transgene, when specifically expressed in the intestine in
mir-83-/- mutant animals, resulted in the restoration of autophagy to wildtype levels, thus, suggesting the cell non-autonomous regulation of autophagy, i.e., inter-tissue/organ autophagic regulation [
87]. These researchers suggested that transportation of
mir-83 across tissue i.e., from intestine to BWM is possibly through extracellular vesicles (EVs).
-
2.
Cell non-autonomous autophagy in systemic body response
Modulation of signaling molecules within one tissue can lead to systemic upregulation/downregulation of autophagy [
4,
87,
88]. This has been elegantly demonstrated recently in
D. melanogaster. D. melanogaster 3rd instar larvae exhibit reduced dTOR activity and a strong autophagy response in their fat body (the mammalian equivalent of the liver) in response to starvation [
89]. The authors demonstrate that activated Dmp53 mediates the reduction of dTOR systemically in response to starvation. Dmp53 mediated systemic reduction of dTOR activity led to reduced dILP2 secretion (
Drosophila Insulin-like peptide 2) and reduced 4E-BP and InR (Insulin Receptor) expression. In
D. melanogaster, dILP2 is secreted from insulin-producing cells (IPCs) located in the brain, and its secretion is stimulated by Upd2 from the body [
89]. The researchers demonstrated that Dmp53 affected dILP2 secretion from IPCs via reduced Upd2 production and secretion from the fat body. Finally, the researchers were able to demonstrate that AMPK mediated Dmp53 activation in the fat body in response to starvation. Thus, these data show that AMPK-dependent Dmp53 activation regulates systemic autophagy cell non-autonomously by modulating Dilp2 secretion from IPCs through Upd2 production and secretion from the fat body [
89] (
Figure 4).
-
3.
Non-autonomously regulated autophagy in Development
During development, autophagy could be non-autonomously regulated. Tissue and organ morphogenesis in animals depends on the balance between cell survival and death. Developmental programmed cell death is mediated by Caspase-dependent apoptosis and context-specific autophagy-dependent cell death. Steroid hormones are critical regulators of physiology and developmental transitions in mammals and insects [
90,
91,
92,
93,
94,
95].
D. melanogaster larval to pupal transition is regulated by Ecdysone, and during this transition, larval tissues are removed, remodelled, and pave the way for the development of adult tissues [
96,
97,
98,
99]. The larval midgut is removed by programmed autophagy. Denton et al. 2019 demonstrated that Dpp (BMP2/4 homolog) signaling needs to be downregulated for midgut clearance by autophagy [
100]. The data show that Dpp transcriptionally inhibits several
Atgs, including
Atg1 and
Atg8a, as well as the
Ecdysone receptor [
100]. In contrast, Ecdysone upregulates several autophagy genes, suggesting a crosstalk between two opposing signaling pathways
viz., Dpp signaling (preventing midgut degradation through downregulating
Atg genes) and Ecdysone signaling (triggers midgut removal through upregulating
Atg genes). Further studies showed that
EcR primary response genes,
BrC and
E74, and several ecdysone biosynthetic genes (spook, phantom, disembodied, shadow, shade) were significantly reduced when the Dpp pathway was upregulated [
100]. Additionally, Denton et al. found that a Dpp-expressing midgut prevents the production of ecdysone in the prothoracic glands (PG) [
100]. Strong nuclear Mad localization was observed in the PG of animals expressing Dpp. This phenotype was not observed in
TkvQ253D, suggesting a signal produced from the midguts, i.e., Dpp, signals to peripheral tissues, including the PG. Thus, Dpp regulates autophagy-dependent midgut removal via cell non-autonomous regulation of ecdysone production, which regulates autophagy cell non-autonomously [
100]
Figure 5A.
Non-autonomous activation of autophagy by membrane-bound proteins can be observed during
D. melanogaster larval salivary gland clearance. Autophagy and Caspases are needed for the degradation of larval salivary glands, but how autophagy regulates the process is poorly understood [
101,
102,
103]. Macroglobulin complement-related (Mcr) is a conserved protein expressed in salivary glands and shows elevated levels after 12 hours of steroid-triggered cell death during puparium formation [
104]. Mcr binding to Draper, an immune receptor and a cell-autonomous autophagy regulator, triggers its phosphorylation by Src42, activating Atg1 [
82,
101,
104,
105]. When Mcr clones are generated, it blocks autophagy in the Mcr mutant cell, but Mcr
+/+ cells neighboring Mcr
-/- cells secrete Mcr in the tissue environment, which leads to activation of autophagy in the Mcr
-/- [
104]
Figure 5B. Mcr also mediates an inflammatory response of macrophages, wherein, secreted Mcr from the wounded epithelial cells activates cell non-autonomous autophagy in macrophages migrating towards the wound [
104]. This report has shown that Mcr is a cell non-autonomous autophagy regulator during programmed cell death (PCD) of larval salivary gland and during epithelial wound healing, but interestingly, nutrient deprivation-induced autophagy of fat body along with the autophagy of dying midgut is not influenced by Mcr, suggesting its localized function [
104].
-
4.
Cell non-autonomous autophagy in Neurodegeneration and protein aggregation pathologies
Amyotrophic lateral sclerosis (ALS) is a lethal neurodegenerative disease wherein motor neurons are lost. ALS is caused by many mutations, of which SOD1 and C9ORF72 are the most common [
106]. It has been reported that piling-up of misfolded ubiquitinated protein aggregates occurs, making this the hallmark of most forms of ALS [
107]. The researchers used motor neurons derived from iPSCs generated from an ALS patient (iPS31c8) and healthy control (iPSC1c1) cultured in Astrocyte Conditioned Medium (ACM) obtained from three ALS patients and three healthy controls, wherein they observed significant viability loss of motor neurons cultured in ALS derived ACM, hinting towards a secretory regulation of motor neurons by astrocytes. Interestingly, the phenotype is unique to motor neurons. Also, several studies show that altered autophagy contributes to ALS pathogenesis. C9ORF72, a mutation that is the most common cause of ALS, modulates the ULK1 activity, thereby regulating autophagy initiation [
108]. HEK293T cells were cultured in ACM obtained from a patient suffering from ALS, and in ACM derived from healthy patient astrocytes, it was observed that LC3-II levels were significantly lower in ALS derived ACM maintained HEK293T cells, implying decreased autophagy. However, no alterations are seen in BEC-1, pULK1, ATG3, or ATG12 in the same cells. In addition to the LC3-II levels, p62 was significantly increased in ALS-derived ACM-maintained cells, implying decreased autophagy [
109]. Thus, it can be inferred that astrocytes of ALS patients secrete factors that target motor neurons wherein these factors impair autophagy in the motor neurons, leading to the accumulation of ALS-related proteins like SOD1 and TDP-43 which otherwise would be degraded, thereby elevating ALS pathogenesis [
109]. When autophagy was induced in the ALS models, it reduced the severity of the disease. p62 puncta in cells treated with control ACM and subjected to rapamycin exposure showed no change. However, ALS ACM-induced p62 puncta return to comparable levels to control ACM-induced cells [
109]. Therefore, it can be inferred from the data provided that autophagy is regulated from outside of the target cell (motor neuron) by the neighboring astrocytes pointing towards non-autonomous regulation of autophagy in ALS pathogenesis (
Figure 6A).
In protein-aggregate pathologies, both cell and cell non-autonomous mechanisms of neurodegeneration are observed, where the latter helps with the propagation of protein aggregates and/or pathologies throughout the nervous system [
110,
111]. In the transthyretin (TTR) systemic amyloidosis, the healthy liver secretes tetrameric TTR into the bloodstream, where TTR dissociates, misfolds, and aggregates, compromising organ systems such as the heart, autonomic nervous system, and peripheral nervous system [
112,
113,
114]. In
C. elegans, TTR toxicity using various mutants of the TTR protein was studied, viz., non-native TTR (NN TTR), WT-TTR, V30M-TTR, and D18G-TTR. In familial amyloid polyneuropathy (FAP) patients, the aggregation of V30M-TTR and its deposition around peripheral and autonomic neurons results in loss of nociception and thermal perception in the extremities [
112]. In
C. elegans, V30M-TTR expression in BWM led to damage to thermal sensory neurons and loss of thermal sensitivity. This effect could be rescued by RNAi treatment against
V30M-TTR in BWM, further supporting cell non-autonomous neuronal proteotoxicity. The authors further provide evidence that TTR tetramers, when secreted into the body-cavity, are then taken up by coelomocytes where TTR is degraded within the lysosomes, i.e., cell non-autonomous degradation of tetramers via autophagosomal-dependent process in coelomocytes [
115] (
Figure 6B). Compromising coelomocytes or their lysosomal system leads to TTR accumulation. This study reveals that coelomocytes trigger a degradative mechanism, which is autophagosome-dependent or analogous. Hence, modulating autophagy flux could regulate TTR tetramer and oligomer levels and ameliorate NN-TTR-mediated neurotoxicity [
115].
-
5.
Cell non-autonomous autophagy in Cancer and supercompetition
Cancerous cells exhibit several metabolic perks, which gives them a growth advantage in stressful tumor niches [
116,
117]. Mitochondrial ATP production is depleted in cancerous cells; however, demand for TCA cycle-derived biosynthetic precursors and NADPH is unaltered or elevated [
118,
119]. Carbon sinking in the TCA cycle comes from glutamine (Gln) to compensate for such conditions. Many cancer cells exhibit higher rates of Gln transport and metabolism due to the high activity of mitochondrial glutaminase [
118,
119,
120,
121,
122]. Gln is sequentially deaminated to Glutamate (Glu) and eventually to α-ketoglutarate (α-KG). A by-product of glutaminolysis is a potentially toxic compound named ammonia, which is excreted out of the cell by diffusion, by transport, or through incorporation into α-keto-acids [
123]. In U2OS cells that are cultured in a medium without replenishment, which are hereafter called cell-conditioned medium (CCM), showed an increase in GFP-LC3 puncta and a decrease in the abundance of p62, both of which are the markers indicative of upregulated autophagy. When the CCM is added to the freshly plated U2OS cells, it induces autophagy [
124,
125]. Through a series of experiments, the authors deduced that CCM has some metabolites/molecules that diffuse within the cells to induce autophagy. This diffusible metabolite was further characterized and was found to be ammonia [
124,
125]. This ammonia was present in mM concentrations both in CCM and CCM-derived distillate and was capable of autophagy induction. Cells directly treated with NH
4OH induced autophagy with similar kinetics, confirming that ammonia induces autophagy [
124,
125]. To validate these findings under
in-vivo conditions, interstitial fluids were collected from human cancer cell line xenografts, and surprisingly, it was found that this fluid has comparable ammonia concentrations to CCM. Expectedly, Gln-free CCM generated in the presence or absence of α-KG lacked autophagy-stimulating activity. In this Gln-free medium, when Gln is added, even then autophagy induction fails to be restored, implying Gln’s does not induce autophagy but acts as a precursor for the generation of ammonia [
124,
125]. Interestingly, adding ammonia to Gln-deprived cells partially restored basal autophagy. The authors also demonstrated that ammonia induces autophagy independently of mTOR [
124,
125]. However, when cells were knocked down for
Ulk1 and were subjected to CCM, even ammonia addition failed to induce autophagy, confirming that ammonia-induced autophagy is mTOR-independent but Ulk1-dependent [
124,
125]. Diffusion of ammonia from tumor cells that utilize glutamine in glutaminolysis provides an autocrine- and paracrine-acting signal that promotes cell-autonomous and non-autonomously regulated autophagy and, in turn, protects cells in different regions of the tumor from stress generated internally as well as environmental stress (
Figure 7C).
An interesting study demonstrated that autophagy could be induced non-autonomously within the tumor microenvironment.
D. melanogaster eye antennal disc (EAD)
RasV12scrib-/- malignant tumor grow and invade the central nervous system and kill the host. Interestingly, as the malignancy progressed, the interaction with the tumor microenvironment triggered autophagy in neighboring non-tumor cells [
126]. This phenotype was not observed in benign Ras
V12 tumors. Interestingly, preventing autophagy within the tumor did not prevent their invasion of the neighboring tissue. However, disruption of autophagy in the microenvironment drastically reduced tumor growth and invasion [
126]. In addition to the local cell non-autonomous autophagy activation, systemic cell non-autonomous autophagy activation is observed for distal organs/ tissues like fat body, midgut, and muscles. The induction of non-autonomously regulated autophagy was abolished by overexpressing activated Yorkie, Unpaired 1, and Unpaired 3 [
126]. (
Figure 7D)
Zinc is a mandatory micronutrient and is involved in a variety of fundamental biological processes [
127,
128,
129,
130,
131]. A study by Wei et al. has shown Zinc and its transporters have a crucial role in tumorigenesis [
132]. The authors used
RasV12 and
Raf GOF Scrib-/- D. melanogaster as models to mimic the behaviours of various human cancers [
133,
134,
135]. Blocking the Zinc transporter (
dZnT7 RNAi) in
RasV12 enhanced the growth of cancer, while in
RafGOF Scrib-/- led to increased growth as well as invasion. Further, they found that JNK was upregulated, and it, in turn, regulated autophagy both autonomously and non-autonomously [
132]. Cell-autonomous autophagy promotes tumor cell growth. Non-cell autonomous autophagy in the surrounding tissues helps the growth of tumors by providing them with vital nutrients [
126,
132,
136]. Chloroquine treatment of
RafGOF Scrib-/- reduced tumor growth and invasion, suggesting that inhibition of autophagy could be a therapeutic strategy for cancers [
132]. dZnT7 knockdown promoter tumor overgrowth and migration through activation of JNK-mediated cell-autonomous and cell non-autonomous autophagy. (
Figure 7D)
The Hippo pathway has been recently shown to be involved in inducing non-autonomously regulated autophagy in eye antennal disc cells in
D. melanogaster. Fat is a core component of the Hippo pathway, and fat mutant cells outgrow their neighboring wild-type cells and develop tumors in adult eyes and heads [
137]. These
fat mutant clones are super competitors (winner), and they induce cell non-autonomous autophagy-mediated cell death in neighboring wild-type loser cells. The loser cells provide nourishment to the winners and aid in the development of the tumor [
137,
138]. Fat mutant clones activate the cell non-autonomous autophagy-mediated cell death in loser cells in an NF-κB-dependent Hid induction. Interestingly, cell non-autonomous autophagy induction depends on Yki-
bantam-mediated TOR signaling activation in winner cells [
137]. TOR signaling activation led to upregulation in protein synthesis in supercompetitors [
137,
138]. It is not clear how autophagy is induced in neighboring cells by fat mutant cells; it may be due to relative differences in protein synthesis levels between winners and losers. A similar phenomenon can be observed even in the competition between Hel25E mutant loser clones when surrounded by wild-type clones [
138]. In this scenario, loser cells have compromised protein synthesis due to a mutation in the ribosomal protein
Hel25E. Thus, it can be speculated that cells with compromised protein synthesis relative to their respective supercompetitors activate autophagy and eventually lead to cell death (
Figure 7A-B). However, how autophagy is induced at the boundary of the winner and loser cells is not understood; but there is a possibility that loser cells respond to some secreted factor form winner cells, which in turn leads to autophagy induction in loser cells [
126,
137,
138]. This data suggests that cell competition can also be autophagy-dependent and may drive tumor expansion during tumorigenesis. This supercompetition is not observed in
RasV12scrib-/- malignant eye antennal disc tumor where cell non-autonomous autophagy is induced in wild-type surrounding cells but does not lead to their cell death; however, it is essential for tumor growth and invasion [
126]. This difference in mode of action needs to be elucidated further (
Figure 7A-D).
-
6.
Cell non-autonomous autophagy in ECM-mediated regulation and Stem cell maintenance
Extracellular matrix (ECM) forms an intricate network and is essential for the transmission of external signals into the cellular interior. Amongst several ECM proteins heparan sulfate-proteoglycans are found both the cell surface and within the ECM and play indispensable roles in various developmental signaling pathways, including WNT/Wg, Hh, TGF-β, and FGF [
139,
140,
141,
142,
143]. In
Drosophila, Heparan sulfate proteoglycans (HSPGs) coordinate interactions between motor neurons and muscle cells via signaling molecules/pathways at the neuro-muscular junctions (NMJs) [
144,
145]. Perturbation of HS polymer synthesis or their modifications in muscle cells, led to changes in mitochondrial numbers, and expression of proteins associated with the Golgi and ER. Intracellular changes were also observed in the subsynaptic reticulum (SSR). [
146]. The SSR is a highly ordered structure where cytoskeletal, neurotransmitter receptor, and signaling molecules are spatially localized and appears to be sensitive to level of HSPGs [
147]. For instance, pancellular or Muscle-specific disruption of HSPG, mediated by
ttvRNAi and
sflRNAi showed ultrastructural changes at the NMJ of motor-neuron terminals, including expansion of spaces between membrane layers of the SSR. Further, TEM analysis showed presence of mitochondria with altered morphology and increase in number of double-membrane vesicles i.e., autophagosomes [
146]. This increase in autophagy in
ttvRNAi and
sflRNAi was confirm by increased Ref(2)P/p62 accumulation, increased number of GFP-Atg8a puncta and Lysotracker Red-positive vesicles [
146]. The authors speculated that HSPG may also be required for autophagy in other tissues/cell types. To test this fatbody specific knockdown of
ttv and
sfl was performed using RNAi and autophagy was assessed in these cells. The data obtained confirmed the observation from muscle cells i.e., increased in GFP-Atg8a puncta and increase in Lysotracker Red positive structures, indicating that loss of HS biosynthesis leads to autophagy induction. The researchers further addressed if HS biosynthesis is required cell-autonomously or cell non-autonomously in fatbody cells. The authors employed sophisticated inducible FLP-FRT based mitotic recombination system in the fat body to generate mutant cells with compromised HS biosynthesis (
ttv-/-; ttv00681/ttv00681) alongside wild-type cells (
ttv00681/+) from heterozygous parental cells.
ttv00681/ttv00681 mutant cell clones in close proximity to cells with uncompromised HS biosynthesis (Wt cells), did not show increased Lysotracker-Red positive puncta, supporting the notion that HS biosynthesis, mainly mediated by
ttv, operates in a non-cell-autonomous manner to regulate autophagy, potentially through modulation of signaling pathways (
Figure 8A).
Cellular homeostasis is important to sustain stem cell pools in different tissues, and autophagy plays a critical role in maintaining homeostatic conditions within these cells [
148,
149]. Germline stem cells are specialized stem cells that produce only eggs or sperm and are crucial for the survival of the species. The authors used
D. melanogaster testis as a model to study the role of autophagy in somatic and germline stem cell maintenance and aging. Basal autophagy was active in somatic cells of the testis, and there were fewer mCherryAtg8a puncta in germ cells than in early cyst cells (CCs) and hub cells [
88]. When genes involved in autophagosome formation like
Atg1, Atg5, and Atg8a were depleted in Cyst Stem Cells (CySCs) and early CCs, it resulted in depletion in CySCs, early CCs, and CCs. It also resulted in a reduction in the GSC number; however, when autophagy is disrupted within GSCs, it did not reduce the GSC number, suggesting that cell-autonomous autophagy is not required in GSCs under homeostatic conditions [
88]. Reduced autophagy in CySCs leads to their loss, suggesting that autophagy in cells is autonomously required [
88].
Further, the authors discovered that
Atg1,
Atg5, and Atg8a knockdown resembled
EGFRDN mutant phenotype, suggesting a relation between EGFR signaling and autophagy [
88,
150,
151]. Disrupting autophagy in CySCs did not lead to significant changes in dpErk intensity, indicating that autophagy does not act upstream of EGFR signaling [
88,
152] However, when
EGFR is activated in early somatic cells, there is a relative increase in autolysosomes, suggesting
EGFR stimulates basal autophagy in somatic cells of the testis. When EGFR signaling is blocked (
EGFRDN or
fos mutants) in early CCs, a significant decrease in mRNA levels of
Atg6 and autolysosome numbers is observed [
88]. Furthermore, Demarco et al., 2020 found that high InR/TOR signaling-induced autophagic suppression is required in the CySC daughter cells to promote differentiation [
88]. Thus, a model can be drawn out from these genetic interactions that EGFR promotes autophagy in CySCs, which helps in maintaining CySCs and GSCs in opposition to InR/TOR signaling, which suppresses autophagy in early CCs and thereby promoting their differentiation [
88] (
Figure 8B). Thus, non-autonomously regulated activation of autophagy is crucial for stem cell maintenance in
D. melanogaster testis.
A recent report on the non-autonomous regulation of autophagy is described in
D. melanogaster male reproductive system. The
D. melanogaster testis harbors GSCs, which give rise to sperm and are regulated by their niche cells called the hub cells and CySCs. BMPs (Dpp and Gbb) are produced by both hub cells and CySCs, and upon binding, BMP receptors in GSCs suppress
bag-of-marbles (
bam), which promotes the differentiation program [
153,
154,
155]. Dpp and Gbb from hub cells and CySCs activate BMP signaling in GSCs and non-autonomously suppress autophagy, which can be detected only at low/basal levels in the GSCs, as seen by reduced Atg5 and Atg8a expression [
156]. Autophagy is upregulated, as seen by increased Atg5 expression and increased mCherryAtg8a puncta formation in the differentiated cells where BMP signaling is inactive and Bam is expressed. Further, the authors showed that upregulation of the Dpp suppressed autophagy in GSCs, as seen by reduced lysotracker staining and reduced number of mCherryAtg8a punctate structures. The authors further demonstrated that Atg5 suppression is mediated by Mad-Med binding to the 5th intronic region of the
Atg5. However, the suppression of Atg8a by BMPs is not at the transcriptional level. Thus, BMPs signaling suppress autophagy in male GSCs cell non-autonomously and via distinct mechanisms [
156].
-
7.
Cell non-autonomous autophagy in mitochondrial degradation
The mitochondrial quality control system governs processes including mitochondrial biogenesis, mitochondrial dynamics, and their degradation [
157,
158,
159]. Mitophagy, the autophagic degradation of dysfunctional mitochondria, is a cell-autonomous process [
157]. Davis et al., identified a phenomenon at the optic nerve head (ONH) of the retinal ganglion cells (RGCs) in wt mice where intact retinal ganglion cell axons shed membrane bound vesicles which were internalized and subsequently degraded in the astrocytes. Using 3D electron microscopic analyses the authors demonstrated that these evulsions in addition to microtubules also contain mitochondria. The authors further provide evidence that these evulsions containing mitochondria are internalized within astrocytes and degraded, as these vesicles colocalized with lysosomes of the astrocytes [
160]. Furthermore, expressing mitochondrial specific dual fluorescent reporter MitoEGFPmCherry only in the RGCs the researchers detected a few mCherry-only expressing (acidified mitochondria) mitochondria in astrocytes supporting the fact that mitochondria of the RGCs are taken up in the astrocytes. In addition, by analyzing colocalization of MitoFISH and TUNEL staining the authors demonstrated that these mitochondria now detected in astrocyte cytoplasm are in early stages of degradation [
160]. These observations support the notion that although axons represent a significant site for mitochondrial degradation in RGCs, not all axonal mitochondria are degraded autonomously within RGC axons as previously believed, but rather by adjacent astrocytes through what the authors term it as transcellular mitophagy or transmitophagy [
160]. Further studies in transmitophagy pathway are essential to understand how this process is controlled at the molecular level.