Mechanistic Target of Rapamycin Signaling Activation Antagonizes Autophagy To Facilitate Zika Virus Replication

The re-emergence of Zika virus (ZIKV) and its association with neurological complications necessitates studies on the molecular mechanisms that regulate ZIKV pathogenesis. The mTOR signaling cascade is tightly regulated and central to normal neuronal development and survival. Disruption of mTOR signaling can result in neurological abnormalities. In the studies reported here, we demonstrate for the first time that ZIKV infection results in activation of both mTORC1 and mTORC2 to promote virus replication. Although autophagy is activated early in infection to counter virus replication, it is subsequently suppressed by mTOR. These results reveal critical roles of mTOR signaling and autophagy in ZIKV infection and point to a possible mechanism underlying ZIKV-induced pathogenesis. Elucidating the role of mTOR signaling in ZIKV infection will provide insights into the mechanisms of ZIKV-induced neurological complications and potential targets for therapeutic approaches.

ABSTRACT

Zika virus (ZIKV), a mosquito-transmitted flavivirus, is linked to microcephaly and other neurological defects in neonates and Guillain-Barré syndrome in adults. The molecular mechanisms regulating ZIKV infection and pathogenic outcomes are incompletely understood. Signaling by the mechanistic (mammalian) target of rapamycin (mTOR) kinase is important for cell survival and proliferation, and viruses are known to hijack this pathway for their replication. Here, we show that in human neuronal precursors and glial cells in culture, ZIKV infection activates both mTOR complex 1 (mTORC1) and mTORC2. Inhibition of mTOR kinase by Torin1 or rapamycin results in reduction in ZIKV protein expression and progeny production. Depletion of Raptor, the defining subunit of mTORC1, by small interfering RNA (siRNA) negatively affects ZIKV protein expression and viral replication. Although depletion of Rictor, the unique subunit of mTORC2, or the mTOR kinase itself also inhibits the viral processes, the extent of inhibition is less pronounced. Autophagy is transiently induced early by ZIKV infection, and impairment of autophagosome elongation by the class III phosphatidylinositol 3-kinase (PI3K) inhibitor 3-methyladenine (3-MA) enhances viral protein accumulation and progeny production. mTOR phosphorylates and inactivates ULK1 (S757) at later stages of ZIKV infection, suggesting a link between autophagy inhibition and mTOR activation by ZIKV. Accordingly, inhibition of ULK1 (by MRT68921) or autophagy (by 3-MA) reversed the effects of mTOR inhibition, leading to increased levels of ZIKV protein expression and progeny production. Our results demonstrate that ZIKV replication requires the activation of both mTORC1 and mTORC2, which negatively regulates autophagy to facilitate ZIKV replication.

IMPORTANCE The re-emergence of Zika virus (ZIKV) and its association with neurological complications necessitates studies on the molecular mechanisms that regulate ZIKV pathogenesis. The mTOR signaling cascade is tightly regulated and central to normal neuronal development and survival. Disruption of mTOR signaling can result in neurological abnormalities. In the studies reported here, we demonstrate for the first time that ZIKV infection results in activation of both mTORC1 and mTORC2 to promote virus replication. Although autophagy is activated early in infection to counter virus replication, it is subsequently suppressed by mTOR. These results reveal critical roles of mTOR signaling and autophagy in ZIKV infection and point to a possible mechanism underlying ZIKV-induced pathogenesis. Elucidating the role of mTOR signaling in ZIKV infection will provide insights into the mechanisms of ZIKV-induced neurological complications and potential targets for therapeutic approaches.

INTRODUCTION

Zika virus (ZIKV) is primarily transmitted by mosquitoes, but other modes of transmission, such as sexual and vertical transmission, have also been reported (1–3). Additionally, virus shedding has been demonstrated in bodily fluids like tears, semen, and urine (4–6). The explosive epidemic in Brazil in 2015-2016 has been associated with serious human diseases such as congenital abnormalities, including fetal growth restriction and microcephaly (7). While the majority of ZIKV infections in adults are asymptomatic and only a fraction of infected individuals may show symptoms of febrile illness, including mild fever, rash, and conjunctivitis (7), ZIKV infection can trigger the development of Guillain-Barré syndrome (8), and reports have demonstrated cell death and reduced proliferation in adult stem cells (9).

ZIKV is a member of the family Flaviviridae. The ∼10.8-kb positive-sense RNA genome of the virus containing a single open reading frame (ORF) is translated into a single polyprotein that is processed by cellular and viral proteases to yield 3 structural proteins, capsid (C), premembrane (prM), and envelope (E), and 7 nonstructural (NS) proteins, NS1, NS2A, NS2B, NS3, NS4A, NS4B, and NS5 (10, 11). While the NS proteins are involved in the replication of the genome, assembly of progeny virions, manipulation of cellular signaling pathways, and evasion of host antiviral responses, the structural proteins are responsible for virus attachment and entry into host cells as well as virus particle formation (10, 12).

Recent studies suggest that ZIKV infection results in defective neuronal cell development and cell death and that these defects might lead to the development of neurological complications (13). Signaling via the mechanistic (mammalian) target of rapamycin (mTOR) is a key pathway in regulating cellular growth, proliferation, and survival (14). mTOR is a serine/threonine kinase belonging to the phosphatidylinositol 3-kinase (PI3K)-related protein kinase family and constitutes the catalytic subunit of two functionally distinct complexes, mTOR complex 1 (mTORC1) and mTORC2 (14). The regulatory associated protein of mTOR (Raptor) and the rapamycin-insensitive companion of mTOR (Rictor) are essential and distinguishing subunits of mTORC1 and mTORC2, respectively. They act as scaffold proteins to help assemble and stabilize the respective complexes (14, 15). Phosphorylation and activation of downstream effectors of mTORC1, such as the p70 ribosomal S6 kinase 1 (p70S6K), the eukaryotic initiation factor 4E (eIF4E)-binding protein 1 (4EBP1), and Unc-51-like autophagy activating kinase 1 (ULK1), regulate cap-dependent mRNA translation, metabolism, and protein/organelle quality control mechanisms (14). mTORC2 phosphorylates several proteins of the AGC family of kinases, such as protein kinase C (PKC), protein kinase B (PKB/Akt), and the serum and glucocorticoid induced kinases (SGK), regulating various aspects of cytoskeletal structure modeling and cellular survival (14).

mTOR signaling is one of the fundamental pathways essential for brain development and is known to regulate autophagy (16). Infection of cells by viruses influences the activities of mTORCs (17–19). While poxviruses disrupt mTORCs’ regulatory cascade (20), adenoviruses activate mTORC1 (21, 22). The NS1 protein of influenza A virus activates mTOR signaling network by phosphorylating and activating Akt, a key protein in the mTOR signaling pathway, and by inhibiting REDD1 (regulated in development and DNA damage responses 1), a key inhibitor of mTORC1 (23). Members of the family Flaviviridae such as West Nile virus (WNV) and dengue virus (DENV) activate PI3K/Akt and mTORC signaling (24, 25), resulting in increased viral protein expression and replication (24). The NS4A and NS4B proteins of ZIKV have been shown to inhibit Akt-mTOR signaling in human fetal neuronal stem cells (26).

Autophagy is a cellular homeostatic process involving the formation of autophagosomes, which engulf protein aggregates, damaged cell organelles, and intracellular pathogens marked for degradation (27). It also plays a major role in eliciting an antiviral response (28). Pathogens like herpes simplex virus 1 (HSV-1) (29), human immunodeficiency virus (HIV) (30), and influenza virus (31) subvert the activation of autophagy to enhance their replication. While induction of autophagy facilitates DENV replication (32), it restricts WNV replication and protein synthesis and acts as an antiviral response of the host cell (33). In contrast, ZIKV infection has been shown to induce autophagy (26). Since mTOR and autophagy are key signaling cascades that regulate many cellular processes, continued efforts on how ZIKV perturbs these pathways are important for understanding of ZIKV infection and pathogenesis.

Here, we demonstrate that ZIKV infection results in the activation of both mTORC1 and mTORC2 in neuronal and glial cells. Inhibition of mTOR kinase reduces ZIKV protein expression and progeny virus production. Additionally, our studies reveal that ZIKV infection induces autophagy at early stages of infection but that later in infection, autophagy is subdued by the concerted activation of both mTORC1 and mTORC2, resulting in viral protein accumulation and virus growth. Our results demonstrate that activation of mTOR signaling and suppression of autophagy are required for ZIKV growth in vitro and provide a framework for further research on the role of these two cellular pathways for understanding of the mechanism(s) underlying ZIKV induced pathogenesis.

RESULTS

ZIKV infection activates mTORC1 and mTORC2 in neuronal and glial cells in culture.

mTOR signaling is known to be modulated by virus infections (17–19). In order to investigate the effect of ZIKV infection on neuronal mTOR signaling, we used neuronal progenitor LUHMES cells (34) and first examined the status of mTORC1 activation in these cells infected with ZIKV by determining the levels of phosphorylation (p) of p70S6K, a downstream substrate of mTORC1 (14, 15). Western blot analysis showed that the levels of p-p70S6K at threonine 389 (T389) increased following ZIKV infection (Fig. 1A). In the LUHMES cells, the phospho-specific antibody used here also detected the phosphorylated second isoform of the protein, p-p85S6K (T412), which is derived from the same gene and contains 23 extra amino acids at the amino terminus (35). The increase in phosphorylation was evident at 12 to 24 h postinfection (hpi) and continued to increase until 48 hpi compared to the mock-infected cells (Fig. 1A). Detectable levels of ZIKV envelope (E) protein were observed at 24 hpi, suggesting that viral protein expression paralleled mTOR activation (Fig. 1A). To determine if mTORC1 activation by ZIKV infection occurs in a cell type-independent manner or is restricted to neuronal precursor cells, we examined the levels of p-p70S6K (T389) in A172 (human glioblastoma) cells infected with ZIKV. Results show that the levels of p-p70S6K (T389) were higher in ZIKV-infected A172 cells (Fig. 1B) than in mock-infected cells, suggesting that mTORC1 activation by ZIKV infection is independent of the type of cells used in these studies.

FIG 1.

ZIKV infection activates both mTORC1 and mTORC2 in neuronal and glial cells in culture. (A) LUHMES cells were either mock infected (M) or infected with ZIKV at an MOI of 1, and cell lysates were prepared at the indicated times postinfection. The lysates were subjected to Western blot analysis to detect p-p70S6K (T389), total p70S6K, ZIKV E protein, and β-actin using the corresponding antibodies. The phospho-specific antibody also detected a slower-migrating p-p85S6K (T412) (top panel). Representative images from three independent experiments are shown. The ratios of p-p70S6K/p70S6K and ZIKV E/β-actin from these images are shown in italics. The p-p70S6K/p70S6K ratio in mock-infected cells and ZIKV E/β-actin ratio in ZIKV-infected culture at 48 hpi were arbitrarily set at 1. Relative electrophoretic mobility of molecular mass markers in kilodaltons is shown on the right. nd, not determined. (B) A172 cells were mock infected (M) or infected with ZIKV at an MOI of 1, and cell lysates were prepared at the indicated times. The lysates were subjected to Western blot analysis to detect p-p70S6K (T389), total p70S6K, pAkt (S473), total Akt, ZIKV E protein, and β-actin using the corresponding antibodies. Representative images from three independent experiments are shown. The p-p70S6K/p70S6K, p-Akt/Akt, and ZIKV E/β-actin ratios from these images are shown in italics. The p-p70S6K/p70S6K and p-Akt/Akt ratios in mock-infected cells and ZIKV E/β-actin ratio in ZIKV-infected culture at 48 hpi were set at 1. The relative electrophoretic mobility of molecular mass markers is shown on the right. nd, not determined.

Since mTORC1 was found to be activated by ZIKV infection, we next examined if mTORC2 is also activated by the virus infection. mTORC2 has distinct substrates and physiological effects (14). Activated mTORC2 phosphorylates the downstream kinase Akt/PKB at serine 473 (S473); therefore, we examined the levels of p-Akt (S473) in A172 cells infected with ZIKV. Our results show a progressive increase in the levels of p-Akt (S473) with time compared to the mock-infected cells (Fig. 1B), suggesting that activation of mTORC2 also occurs in ZIKV-infected cells. Thus, our results presented here demonstrate for the first time that ZIKV infection activates both mTORC1 and mTORC2 in different cell types, including neuronal and glial cells in culture, and that this effect parallels ZIKV protein expression.

ZIKV replication is dependent on mTORC1 and mTORC2 activity.

Enhanced mTOR activity has been linked to both increased and decreased viral replication (18). Since our results suggested that ZIKV infection activates mTOR, we next investigated whether activation of mTORCs is required for ZIKV replication. Torin1 potently and selectively inhibits mTOR kinase activity by an ATP-competitive mechanism (36). To examine the effect of mTOR kinase inhibition on ZIKV replication, cells were pretreated with or without Torin1 for 1 h, infected with ZIKV, and subsequently incubated in the presence of Torin1 for 48 h. Western blot analysis of infected cell extracts showed that while the levels of p-p70S6K (T389) were considerably higher in ZIKV-infected cells than in mock-infected cells, the presence of Torin1 led to nearly undetectable levels of p-p70S6K (T389) in the virus-infected cells (Fig. 2A). Likewise, levels of p-Akt (S473), which were higher in ZIKV-infected cells, were undetectable in the presence of Torin1 (Fig. 2A). Importantly, ZIKV E protein expression was greatly reduced in the presence of Torin1 (Fig. 2A). Examination of infectious virus yield from the infected cells revealed significantly lower virus titers in cells treated with Torin1 at 48 hpi and 72 hpi than in untreated cells (Fig. 2B). These results demonstrate that mTOR kinase activity is essential for viral protein expression and progeny production. Although the virus growth at 24 hpi and 96 hpi was less in the presence of Torin1 than in its absence, the inhibitory effect was not significant. The observed lack of significant inhibitory effect of Torin1 at these time points could be due to the fact that ZIKV growth is not substantial at 24 hpi, as we have seen previously (37), or to a decrease in the viability of cells at 96 hpi because of virus-induced cell death. Indeed, viability of ZIKV-infected cells was significantly reduced at 96 hpi, and the presence of Torin1 reversed ZIKV-induced cell death (Fig. 2C). These results clearly reveal an essential role for the mTOR activity in viral protein expression, infectious-progeny production, and ZIKV-induced cell death.

FIG 2.

ZIKV replication requires both mTORC1 and mTORC2 activity. (A) A172 cells were either left untreated (−) or pretreated (+) with Torin1 for 1 h prior to being mock infected or infected with ZIKV at an MOI of 1 in the continued presence of Torin1. Cell lysates were prepared at 48 hpi and subjected to Western blot analyses to detect p-p70S6K (T389), total p70S6K, pAkt (S473), total Akt, ZIKV E protein, and β-actin using the corresponding antibodies. Representative images from three independent experiments are shown. The p-p70S6K/p70S6K, p-Akt/Akt, and ZIKV E/β-actin ratios from these images are shown in italics. The p-p70S6K/p70S6K and p-Akt/Akt ratios in mock-infected, non-Torin1-treated cells and ZIKV E/β-actin ratio in ZIKV-infected, non=Torin1-treated culture were arbitrarily set at 1. Relative electrophoretic mobility of molecular mass markers is shown on the right. nd, not determined. (B) Cell culture supernatants from untreated (vehicle) or Torin1-treated and ZIKV-infected cells were collected every 24 hpi for quantitation of virus yield by plaque assay. Data from four independent experiments are shown. (C) Cell viability was evaluated by flow cytometry in mock-infected (mock) and ZIKV-infected (ZIKV) cells at 96 hpi in the presence or absence (vehicle) of Torin1. (D) The experiment was done as for panel A, except that the cells were pretreated without (−) or with (+) rapamycin 1 h prior to infection and continued in the presence of the drug. Representative images from three independent experiments are shown. The p-p70S6K/p70S6K, p-Akt/Akt, and ZIKV E/β-actin ratios from these images are shown in italics. The p-p70S6K/p70S6K and p-Akt/Akt ratios in mock-infected, non-rapamycin-treated cells and ZIKV E/β-actin ratio in ZIKV-infected, non-rapamycin-treated cultures were arbitrarily set at 1. nd, not determined. (E) Virus yield in the absence (vehicle) or presence of rapamycin as described for panel B. Data from four independent experiments are shown. (F) Effect of rapamycin on ZIKV-induced cell viability as described for panel C. Data from four independent experiments are shown. Error bars represent SEM. ns, nonsignificant; ****, P ≤ 0.0001.

To determine if ZIKV replication is dependent on either mTORC1, mTORC2, or both, we used rapamycin. Rapamycin has been reported to complex with FK506-binding protein (FKBP), a nonobligate component of mTORC1. The FKBP-rapamycin complex interacts with mTOR kinase and inhibits its activity (38, 39). A172 cells were pretreated with rapamycin for 1 h, infected with ZIKV, and subsequently incubated in the presence of rapamycin for 48 h. We observed that ZIKV-induced p-p70S6K (T389) was considerably reduced in the presence of rapamycin (Fig. 2D, lane 4) along with reduction in the levels of the viral E protein. While rapamycin is known to primarily inhibit mTORC1, resulting in inhibition of p70S6K (T389) phosphorylation, it can also suppress the assembly and function of mTORC2, leading to inhibition of Akt phosphorylation at S473 (40). Accordingly, the levels of p-Akt (S473) were also reduced in the presence rapamycin (Fig. 2D, lane 4), indicating that mTORC2 activity induced by ZIKV infection is also inhibited by rapamycin. Similar to what was observed in the presence of Torin1, culture supernatants showed significantly reduced virus titers from infected cells treated with rapamycin at 48 hpi and 72 hpi (Fig. 2E). Again, the reduced virus yield and observed lack of significant difference in virus titers at 96 hpi is due to increased cell death at 96 hpi, which was prevented by rapamycin (Fig. 2F). Overall, our results demonstrate that activation of mTOR is required for ZIKV protein expression and virus growth. However, the pharmacological inhibition of mTOR signaling was inconclusive with regard to whether activation of either mTORC1 or mTORC2 is required for virus replication.

Both mTORC1 and mTORC2 are required for ZIKV replication.

To unequivocally examine the role of mTOR in ZIKV replication, we conducted experiments in which critical protein components of the mTORC1, mTORC2, or both were depleted. Since the mTOR kinase is the central component of both complexes, whereas Raptor and Rictor are the unique scaffolding subunits of mTORC1 and mTORC2, respectively, we used specific small interfering RNA (siRNA)-mediated depletion of these proteins to downregulate their activities and examine the effects on ZIKV E protein expression and progeny production. In cells transfected with siRNA targeting mTOR kinase, the levels of mTOR protein were greatly reduced compared to that seen with nontargeting (NT) siRNA (Fig. 3A, lanes 3 and 4). Similarly, when siRNAs targeting Raptor (Fig. 3A, lanes 7 and 8) or Rictor (lanes 5 and 6) were used, depletion of the corresponding proteins was also observed. Importantly, the reduction in these protein levels was associated with downregulation of phosphorylation of the corresponding substrates: p70S6K for mTORC1 (Raptor) (Fig. 3B, lanes 7 and 8) and Akt for mTORC2 (Rictor) (Fig. 3B, lanes 5 and 6). These results corroborate that siRNA-mediated depletion of Raptor or Rictor leads to significant downregulation of the activities of mTORC1 or mTORC2.

FIG 3.

mTOR complex 1 (mTORC1) and mTORC2 regulate ZIKV replication. (A) A172 cells were transfected with either nontargeting (NT) siRNA or siRNAs targeting mTOR kinase (mTOR), Rictor, or Raptor for 72 h. The cells were subsequently mock infected (−) or infected with ZIKV (+) at an MOI of 1. Cell lysates were prepared at 48 hpi and subjected to Western blot analyses to detect mTOR, Raptor, Rictor, ZIKV E protein, and β-actin using the corresponding antibodies. Representative images from three independent experiments are shown. The mTOR/β-actin, Rictor/β-actin, Raptor/β-actin, and ZIKV E/β-actin ratios from these images are shown in italics. The mTOR/β-actin, Rictor/β-actin, and Raptor/β-actin ratios from uninfected and NT siRNA-treated cells were set at 1, as was the ZIKV E/β-actin ratio in ZIKV-infected and NT siRNA-treated cells. Relative electrophoretic mobility of molecular mass markers is shown on the right. (B) Cell lysates from the experiment whose results are shown in panel A were subjected to Western blot analysis to detect p-p70S6K (T389), total p70S6K, pAkt (S473), total Akt, and β-actin using the corresponding antibodies. Representative images from three independent experiments are shown. The p-p70S6K/p70S6K and p-Akt/Akt ratios from these images are shown in italics. (C) Infectious-virus production in cells depleted of mTOR, Rictor, or Raptor. The experiments were conducted as described for panel A, and culture supernatants from infected cells were collected at various times and assayed for infectious virus by plaque assay. Data from three independent experiments are shown. The dashed line represents the limit of detection. #, virus titer was below the limit of detection. Error bars represent SEM. a, two-way ANOVA, P ≤ 0.0001; ****, P ≤ 0.0001.

In cells infected with ZIKV, a dramatic reduction in the levels of E protein was observed when mTOR kinase was depleted (Fig. 3A, compare lane 4 with lane 2), consistent with the results in Fig. 2 obtained with the use of Torin1 and rapamycin. Importantly, when Raptor (Fig. 3A, lane 8) or Rictor (lane 6) was depleted from the cells through the use of the corresponding siRNAs, the levels of the viral E protein in these infected cells were also dramatically reduced. These results suggest that both mTORC1 and mTORC2 are necessary for optimal levels of viral protein expression. It is interesting that while depletion of Raptor led to nearly undetectable levels of E protein expression (Fig. 3A, lane 8), Rictor depletion led to detectable but significantly reduced (lane 6) levels of E expression. These results show that the requirement for mTORC1 in viral E protein expression is more important than the requirement for mTORC2. However, mTORC2 is also a critical regulator of ZIKV replication.

Consistent with drastic reduction in viral E protein expression, we also observed significant reduction in infectious-progeny production in cells with depleted mTOR kinase, Raptor, or Rictor (Fig. 3C). Although depletion of mTOR kinase or Rictor led to levels of production of infectious progeny that were significantly lower than those in cells treated with NT siRNA, the effect of Raptor depletion was most dramatic in that infectious-virus production was below the level of detection at 24, 48, or 72 hpi, whereas at 96 hpi, the virus titers were significantly lower than those from NT siRNA-treated cells (Fig. 3C). Thus, our results show that both mTORC1 and mTORC2 are required for ZIKV protein expression and infectious virus production; however, mTORC1 appears to play a more important role in these processes.

ZIKV infection induces autophagy, which suppresses the virus replication.

As mTOR signaling is known to regulate autophagy (16), a cellular homeostatic pathway that can function as an antiviral cellular process (28), we investigated the role of autophagy in ZIKV infection. Flaviviruses, including ZIKV, have been reported to activate autophagy upon infection (26, 41). Microtubule-associated protein 1A/1B-light chain 3 (LC3) is a soluble protein that is recruited to autophagosomal membranes and is commonly used as a marker for changes in autophagosome dynamics. Therefore, we examined the distribution of LC3 in cells infected with ZIKV at various times postinfection.

Human neuroblastoma SK-N-SH cells stably expressing LC3 labeled with enhanced green fluorescent protein (EGFP-LC3) (42) were mock infected or infected with ZIKV. Following incubation for different lengths of time, the cells were either untreated or treated with chloroquine (CQ) 4 h prior to the termination of the experiment to avoid any unintended negative effects of CQ on ZIKV infection (43). The cells were subsequently stained with LysoTracker Red (a marker for lysosomes) and examined by fluorescence microscopy. Results show that ZIKV infection significantly increased the accumulation of EGFP-LC3 punctum clusters at 12 and 24 hpi compared to mock-infected cells in the presence or absence of CQ, an inhibitor of lysosomal cargo degradation (Fig. 4A and B). These results suggest that ZIKV infection increases autophagy flux at early time points of infection.

FIG 4.

Autophagy is induced in cells infected with ZIKV. (A) Cells stably expressing EGFP-LC3 were mock infected or infected with ZIKV at an MOI of 1. The cells were left untreated or treated with CQ for 4 h prior to the termination of the experiment and also stained with LysoTracker Red before being examined by fluorescence microscopy. Arrowheads in merged images indicate EGFP-LC3 punctum clusters. Arrows indicate enlarged cells with a higher abundance of EGFP-LC3 clusters. Bar, 10 μm. (B) Quantification of EGFP-LC3 puncta per cell (n ≥ 37) from three representative fields. Error bars represent SEM. ***, P ≤ 0.001 (one-way ANOVA on ranks).

To corroborate these observations, we evaluated changes in the levels of lipidated LC3-II and p62, an autophagy receptor with the ability to coaggregate in the presence of ubiquitinated substrates (44). A172 cells infected with ZIKV showed an early increase in the levels of p62 (monomers and aggregates) at 12 and 24 hpi in comparison to mock-infected cells (Fig. 5A). Interestingly, the p62 levels decreased over time, and by 48 hpi, they were comparable to those in mock-infected cells (Fig. 5A). The levels of p62 aggregates also closely paralleled those of monomeric p62 at various times after ZIKV infection. Similarly, an increase in LC3-II level was observed between 12 and 24 hpi, which decreased afterward (Fig. 5A). These results show that ZIKV infection results in activation of autophagy at early times after infection and is suppressed as infection proceeds. In the presence of CQ, the levels of p62 and LC3-II increased slightly further in the infected cells at 12 hpi (Fig. 5B, lane 6) in comparison to mock-infected cells (lane 2) and tapered off at later times postinfection. These results corroborate that ZIKV infection induces an increase in autophagy flux at early times (12 to 24 h) following infection.

FIG 5.

ZIKV-induced autophagy suppresses virus replication. (A) A172 cells were either mock infected (M) or infected with ZIKV at an MOI of 1. Cell lysates were prepared at various times postinfection and subjected to Western blot analysis to detect p62 levels. Representative images from three independent experiments are shown. Monomeric p62 levels are presented at the top. A longer exposure of the same blot showing the p62 monomer as well as the p62 aggregates toward the top of the blot is presented in the panel below that. Representative data from three experiments showing the levels of LC3-II in ZIKV-infected cells at various times postinfection are also shown. The p62 monomer/β-actin, p62 aggregates/β-actin, and LC3-II/β-actin ratios are shown in italics, with the ratios in mock-infected culture set arbitrarily at 1. Relative electrophoretic mobility of molecular mass markers is shown on the right. (B) A172 cells were either mock infected (M) or infected with ZIKV at an MOI of 1. CQ was added 4 h prior to harvesting of cell lysates at the indicated time points. Levels of p62, LC3-II, and β-actin were analyzed by Western blotting. Representative images from three independent experiments are shown. The p62/β-actin, and LC3-II/β-actin ratios are shown in italics, with the ratios in mock-infected culture set arbitrarily at 1. (C) A172 cells were mock infected (M) or infected with ZIKV at an MOI of 1 in the presence (+) or absence (−) of 3-MA, cell lysates were prepared at the indicated time points, and the levels of p62, LC3-II, viral E protein, and β-actin were examined by Western blotting. Representative images from three independent experiments are shown. The p62/β-actin, LC3-II/β-actin, and ZIKV E/β-actin ratios are shown in italics, with the ratios set arbitrarily at 1 as shown. (D) Titers of infectious progeny virus, determined by plaque assay from ZIKV-infected cells treated with 3-MA. Data from four independent experiments are shown. Error bars represent SEM. *, P ≤ 0.05; ****, P ≤ 0.0001.

To assess the effect of autophagy induction on viral protein expression, we examined the levels of E protein in cells infected with ZIKV in the presence or absence of 3-methyladenine (3-MA), a purine analog known to block autophagosome elongation by inhibiting the class III PI3K signaling cascade (45, 46). Inhibition of autophagy with 3-MA led to the accumulation of p62 that was paralleled by a decrease in the levels of lipidated LC3-II in infected and mock-infected cells (Fig. 5C). Importantly, we observed an increase in the viral E protein accumulation in infected cells treated with 3-MA compared to cells not treated with 3-MA (Fig. 5C), indicating that autophagy promoted the degradation of viral proteins. Additionally, infectious virus production was significantly enhanced in cells treated with 3-MA compared to untreated cells (Fig. 5D). Taken together, these results show that early induction of autophagy upon ZIKV infection (12 to 24 hpi) is indeed a cellular defense mechanism that suppresses viral protein accumulation and progeny production, but at later stages in ZIKV infection (≥24 hpi), autophagy is suppressed to allow ZIKV protein accumulation and progeny production.

mTOR activation suppresses autophagy and allows ZIKV replication.

We next wanted to investigate if the role of mTOR in the regulation of viral protein expression and progeny production is linked to autophagy. mTORC1 has been shown to negatively regulate autophagy by phosphorylating ULK1 at serine 757 (S757) (47). To this end, we examined lysates of ZIKV-infected cells harvested at different times following infection for p-ULK1 (S757) by mTORC1. Our results show an increase in the levels of p-ULK1 (S757) with time after the virus infection (Fig. 6A) that parallels the gradual decline in the levels of p62 and LC3-II (Fig. 5A), suggesting that ZIKV-induced phosphorylation of ULK1 suppresses autophagy. Accordingly, inhibition of mTORC1 by Torin1 ablated ULK1 phosphorylation (Fig. 6B) with a concomitant increase in autophagy, as evidenced by the enhanced degradation of p62 and LC3-II (Fig. 6B). As shown before, the level of ZIKV E protein was reduced substantially in the presence of Torin1 (Fig. 6B). These results suggest that during ZIKV infection, phosphorylation and inhibition of ULK1 by mTOR inhibit autophagy and facilitate ZIKV E protein accumulation.

FIG 6.

mTOR activation suppresses autophagy through ULK1 phosphorylation and allows ZIKV replication. (A) Phosphorylation of ULK1 is enhanced in ZIKV-infected cells. A172 cells were either mock infected (M) or infected with ZIKV at an MOI of 1. Cell lysates were prepared at various times postinfection and subjected to Western blot analysis for p-ULK (S757), total ULK1, and β-actin. Representative images from three independent experiments are shown. The p-ULK1/ULK1 ratio is shown in italics, with the ratio in mock-infected culture set arbitrarily at 1. (B) Inhibition of mTORC1 by Torin1 leads to reduced ULK1 phosphorylation and reduced detection of the viral E protein. A172 cells treated with (+) or without (−) Torin1 were mock infected (M) or infected with ZIKV as for panel A. Cell extracts were prepared, and the levels of p-ULK1 (S757), total ULK, p62, LC3-II, ZIKV E protein, and β-actin were examined by Western blotting. Representative images from three independent experiments are shown. The p-ULK1/ULK1, p62/β-actin, LC3-II/β-actin, and ZIKV E/β-actin ratios are shown in italics, with the ratios set arbitrarily at 1 as shown. (C and D) Inhibiting autophagy under conditions of mTORC inhibition rescues viral E protein accumulation (C) and progeny production (D). A172 cells were either mock infected (−) or infected with ZIKV (+) as for panel A and treated without (−) or with (+) various inhibitors as shown at the top of panel C. Cell extracts were prepared at 48 hpi and analyzed by Western blotting for p62, LC3-II, the viral E protein, and β-actin. Representative images from three independent experiments are shown. The p62/β-actin, LC3-II/β-actin, and ZIKV E/β-actin ratios are shown in italics, with the ratios set arbitrarily at 1 as shown. Under similar experimental conditions, culture supernatants from the infected cells at 72 hpi were assayed for infectious progeny production (D) by plaque assay. Data in panel D are from three independent experiments. Error bars represent SEM. ***, P ≤ 0.001.

Since inhibiting mTOR kinase resulted in increased autophagy with a concomitant decrease in viral E protein levels, we wanted to investigate if inhibiting autophagy with 3-MA (class III PI3K inhibitor) or MRT6891 (ULK1 inhibitor) under conditions in which mTOR kinase activity has already been impaired would result in rescue of viral E protein levels and infectious-progeny production. Data from such an experiment would provide unequivocal evidence for a direct link between mTOR activation, suppression of autophagy, and viral replication. The enhanced autophagy induced by Torin1, as evidenced by the increased degradation of p62 and LC3-II, was reversed by 3-MA (Fig. 6C, compare lanes 3 and 4 with lanes 9 and 10) or MRT68921 (Fig. 6C, compare lanes 3 and 4 with lanes 11 and 12) in cells infected with ZIKV or left uninfected. However, only in cells infected with ZIKV did 3-MA and MRT68921 prevent the decrease in E protein levels induced by Torin1 (Fig. 6C, compare lane 4 with 10 and lane 4 with 12). Furthermore, infectious virus production, which is also inhibited significantly in the presence of Torin1, could be rescued to the untreated-control levels in the presence of 3-MA or MRT68921 (Fig. 6D). Overall, we conclude that activation of mTOR in ZIKV infected cells suppresses autophagy to allow virus protein expression and progeny production.

DISCUSSION

Studies have demonstrated that ZIKV infection hinders growth and development of neurospheres and organoid cultures in vitro (48–50) and also causes abnormal neurological development and death in fetal mice (51–53). However, the molecular mechanisms of ZIKV-induced pathogenesis are incompletely understood. mTOR signaling is critical for many cellular processes, such as growth, survival, and proliferation (14). Many viruses are known to either activate or suppress mTORC signaling to facilitate viral replication (18). Additionally, prototypic flaviviruses such as WNV, DENV, Japanese encephalitis virus (JEV) and hepatitis C virus (HCV) activate the Akt/mTOR pathway and this activation may play a proviral or an antiviral role (18). While inhibition of mTOR activates autophagy and enhances DENV replication (19), mTORC1 supports WNV replication (24). With regard to HCV, the role of mTOR appears to be controversial in that some studies suggest a proviral role of mTOR activation (54, 55), while others suggest an antiviral role (56–58). Additionally, activation of mTOR and mTOR-dependent mechanisms is implicated in the enzootic transmission cycle of some flaviviruses involving the mosquito vector (59). Therefore, modulation of mTOR signaling appears to be integral to flavivirus transmission, replication, and pathogenesis.

Using human neuronal progenitor (LUHMES) and glioblastoma (A172) cells, we observed activation of mTORC1 following infection with ZIKV. Further studies revealed that ZIKV infection activated not only mTORC1 but also mTORC2 to facilitate viral-protein expression, accumulation, and progeny virus production. Our studies also revealed that early during ZIKV infection, activation of autophagy restricts viral protein accumulation to undetectable levels, but at later times, autophagy is suppressed by the inhibitory phosphorylation of ULK1 via activation of mTOR, leading to increased levels of viral protein accumulation and progeny production.

ZIKV infection induces autophagy (60–62), but there are conflicting results as to the effect of autophagy on ZIKV infection. Some studies report that inhibition of autophagy reduces viral replication and limits viral transmission (60, 62, 63). ZIKV-induced autophagy was also shown to be mediated by inhibition of the Akt-mTOR pathway by the viral NS4A and NS4B proteins, leading to defective neurogenesis (26). In contrast, other studies have reported that autophagy facilitates virus clearance in phagocytes (64) and protects against ZIKV infection (65, 66). On the other hand, pharmacological and genetic inhibition of autophagy was shown to have no adverse effect on ZIKV replication in glial cells (67). While the conflicting findings with regard to the role of autophagy and mTOR in ZIKV infection could be attributed to the different cell types and experimental model systems tested, they could also be linked to the temporality of the events being analyzed. In addition, the experimental tools used to study mTOR signaling might be a source of discrepancy. While phosphorylation of mTOR at S2448 has been used in many studies as a marker for mTOR activation, there is clear evidence that pS2448 does not correlate with mTOR activity (68). Using different cell types, we demonstrated that mTORC1 and mTORC2 are consistently activated by evaluating the phosphorylation status of their direct substrates, p70S6K, ULK1 (mTORC1), and Akt (mTORC2), which constitute more reliable readouts of mTOR activation (14). We have clearly demonstrated that while autophagy is initiated early during the infection process, it is suppressed later on by ZIKV-induced mTOR signaling. Very few studies have addressed the time course of changes in such events, highlighting the importance of our results.

The role of the individual complexes in ZIKV protein expression and progeny production could not be ascertained using Torin1 or rapamycin due to the lack of specificity of these compounds. By using siRNAs to deplete the critical components of the two complexes, we were able to firmly conclude that both mTORC1 and mTORC2 are required for efficient ZIKV replication, although a differential effect was noted in their requirements. mTORC1 appears to play a more dominant role in ZIKV replication than mTORC2. This is borne out from our observations that Raptor depletion led to extremely low to undetectable levels of viral protein expression and infectious progeny production (Fig. 3), while Rictor depletion led to significantly reduced but still detectable levels of E protein expression and progeny production.

Several studies have provided evidence for a major role for mTORC1 in virus replicative processes, but the role for mTORC2 has been less clear. Replication of WNV, HCV, influenza virus, Andes virus, and herpesviruses require mTORC1 but not mTORC2 (23, 24, 55, 69–71). Only one previous study showed that Rictor (mTORC2) is primarily involved in HCMV replication (72). Although the NS5A protein of HCV activates mTORC2 (73), no experimental evidence links this activation to viral replication. Therefore, one important finding from our studies reported here is the significant involvement of mTORC2 in replication of ZIKV. mTORC2 may play an indirect role, likely upstream of mTORC1 activation through phosphorylation of Akt, an activator of mTORC1 (74). Since a negative feedback loop between mTORC1 activation and mTORC2 has been described, in which p70S6K phosphorylates Rictor and decreases mTORC2 function (75, 76), it is possible that ZIKV also overcomes this negative feedback loop to facilitate the potentiation of mTORC1 activation by mTORC2.

It is interesting that depletion of mTOR kinase itself resulted in still-detectable levels of E protein expression and progeny production, while knockdown of Raptor exerted a more robust negative effect. Knockdown of Rictor induced a more pronounced decrease in Akt phosphorylation than mTOR downregulation (Fig. 3). Since mTOR kinase is central to both mTORC1 and mTORC2, our expectation was that depletion of mTOR would have the most significant negative impact on viral E protein expression and progeny production as well as on the corresponding downstream substrates, compared to the Raptor or Rictor depletion. A direct comparison between the knockdown of mTOR, Rictor, and Raptor cannot be made by just comparing the changes in protein levels or the phosphorylation of the corresponding substrates, as this would require that the antibodies used for detection have similar signal-to-noise ratio efficiencies. However, the possibility exists that Raptor might partially regulate ZIKV replication independently of mTOR. Indeed, a recent study reports mTOR-independent functions for Raptor (77). Similarly, Rictor has been shown to form mTOR-less complexes that mediate different processes, including Akt phosphorylation (78, 79). Addressing the possibility of an mTOR-independent role of Raptor and Rictor in ZIKV infection will be an interesting area of research that requires further experimentation.

Our inability to detect viral proteins in ZIKV-infected cells prior to 24 hpi was surprising given that the viral genome is a competent template for translation immediately after entry into cells. We considered the possibility that autophagy was activated early in ZIKV-infected cells, resulting in inhibition of viral protein accumulation to undetectable levels. Flaviviruses are known to activate autophagy, which plays an important role in replication and/or pathogenesis (41). For example, the viral NS4A-induced autophagy protects DENV- and Modoc virus (a murine flavivirus)-infected cells and enhances virus replication (80), whereas ZIKV NS4A- and NS4B-induced autophagy inhibits neurogenesis (26). Autophagy is also induced in DENV- and HCV-infected cells to support viral replication (32, 81), while in a separate study, autophagy was found to play no role in WNV infection (82). Our results (Fig. 5) show that autophagy was indeed induced early (within 24 hpi) in ZIKV-infected cells and that thereafter, there was a gradual decline in the autophagy flux. The decline in autophagy flux was concomitant with an increase in mTOR activation, as seen by increased phosphorylation of ULK1. In addition to p-ULK1, mTORC1 activation has been shown to inhibit autophagy at other signaling steps, so we cannot rule out other targets that mediate the negative regulation of autophagy by ZIKV-induced mTORC1 activation (83). Based on our findings reported here, we propose that activation of autophagy is an early antiviral response against ZIKV infection (Fig. 7A) that is then suppressed by the virus-induced activation of mTORC signaling, leading to enhanced viral protein accumulation and progeny production (Fig. 7B). However, it is important to recognize that the temporality of these events might change depending on the cell types or experimental model systems used and the conditions of infection, including virus dose and length of infection. While inhibition of autophagy increased viral protein accumulation, it occurred only at 24 hpi, which suggests other unexplored mechanisms by which the cell limits viral replication at early time points of infection.

FIG 7.

Scheme of ZIKV-induced regulation of autophagy, mTOR signaling, and effects on viral protein accumulation, progeny virus production and cell death in infected cells. (A) ZIKV infection induces an early activation of autophagy (≤24 hpi) that mediates the degradation of viral proteins, preventing viral assembly. (B) Later stages of infection (≥24 hpi) activate mTORC1 and mTORC2 to mediate the phosphorylation and inhibition of ULK1, whose activity is necessary for autophagy initiation. Inhibition of autophagy facilitates accumulation of viral proteins, infectious progeny release, and cell death. The time frame of the events delineated here is likely dependent on the cell type used for infection, the exposure time, and dosage of the virus. See Discussion for more details.

The mechanisms involved in autophagy activation by ZIKV are incompletely understood. Although NS4A and NS4B proteins have been shown to induce autophagy through inhibition of mTOR signaling (26), we have observed that the induction of autophagy precedes viral protein accumulation with no evidence of inhibition of mTOR signaling. A recent report demonstrated that ZIKV induces an NF-κB-dependent autophagy pathway mediated by STING (65). Another report demonstrated that activation of NF-κB and autophagy by ZIKV in astrocytes seems to depend on TL3 receptors (67). Thus, autophagy could be triggered by the induction of antiviral innate immune responses prior to viral replication. It is also possible that AMP-activated kinase (AMPK), which is activated by ZIKV (84), could also induce autophagy. It is unknown at this time what viral factors drive activation of autophagy early in infection and mTORC signaling. However, we have observed that autophagy is not activated when UV-inactivated ZIKV is used in the experiments (data not shown), suggesting that live virus infection is necessary. Which antiviral immune responses and viral structural and/or nonstructural proteins are involved in the activation of autophagy and the later induction of mTOR signaling is not known at this time and remains to be investigated. Bystander effects from infected cells could also contribute to the modulation of autophagy. Indeed, previous reports have demonstrated that the release of soluble factors, including cytokines, can regulate autophagy in noninfected cells (85–87).

Since hyperactive mTOR signaling is linked to exhaustion of stem cell niches and loss of mTOR activity leads to depletion of progenitor cells responsible for the normal development of the brain (88), our results might also point to a potential mechanism(s) of ZIKV-induced neurological deficits in developing fetal brains.

MATERIALS AND METHODS

Cells and viruses.

Lund human mesencephalic (LUHMES) cells (ATCC CRL-2927), a subclone of tetracycline-controlled, v-myc-overexpressing human mesencephalic-derived cell line MESC2.10 (34), obtained from the American Type Culture Collection Biosource Center, were cultured as described previously (34). Briefly, cultureware was precoated with 50 μg/ml poly-l-ornithine (Sigma-Aldrich; catalog no. P2533-10G) and 1 μg/ml fibronectin (BD Biosciences; catalog no. 354008). Cells were cultured in Dulbecco’s modified Eagle’s medium (DMEM; Gibco; catalog no. 12800-082) supplemented with Neuroplex N-2 (Gemini; catalog no. 400-163), 2 mM l-glutamine (HyClone; catalog no. SH3003401), and 40 ng/ml recombinant basic fibroblast growth factor (Peprotech; catalog no.100-18C). A172 (ATCC CRL-1620) cells were grown in DMEM supplemented with 10% heat-inactivated fetal bovine serum (FBS) and 1× antibiotics (100 U/ml penicillin and 20 U/ml streptomycin). The stable human neuroblastoma SK-N-SH cells expressing EGFP-LC3 (42) were grown in DMEM containing 10% FBS and 1× antibiotics. Cells were maintained at 37°C and 5% CO₂ atmosphere in a humidified incubator. Vero cells (ATCC CRL-1586) were grown in DMEM supplemented with 10% FBS and 1× antibiotics.

For pharmacological inhibition of mTOR kinase, cells were either left untreated or pretreated with 250 nM Torin1 or 1 μM rapamycin for 1 h prior to being mock-infected or infected with ZIKV at a multiplicity of infection (MOI) of 1 with or without the inhibitors and incubated in medium with or without the inhibitors. Similarly, for the inhibition of autophagy, cells were either left untreated or treated with 3-MA (5 mM) or MRT68921 (1 μM) for 1 h prior to infection with ZIKV with or without the inhibitors and further incubated in medium with or without the inhibitors until the time points indicated in the legend to Fig. 6C and D. For the inhibition of autophagic flux, cells were treated with CQ (50 μM) 4 h prior to the termination of the experiment. For virus yield, 100 μl of culture supernatant was collected at the time postinfection indicated in the legend to Fig. 6D, and infectious virus titers were determined by plaque assay (89). Stocks of the infectious clone derived from MR766 ZIKV were prepared in Vero cells as described previously (37).

Antibodies and other reagents.

Anti-phosphorylated (p)-p70S6K (T389)/(p)-p85S6K (T412) (catalog no. 9234S), anti-p70S6K (catalog no. 2708S), anti-p-Akt (S473) (catalog no. 9018B), anti-Akt (catalog no. 2938P), anti-mTOR (catalog no. 2983), anti-Rictor (catalog no. 2114), and anti-Raptor (catalog no. 2280) antibodies were obtained from Cell Signaling Technologies. Anti-p62 (catalog no. ab109012) was obtained from Abcam. Anti-ZIKV E antibody was obtained from GeneTex (catalog no. GTX1333325). Anti-β-actin (catalog no. A2228), horseradish peroxidase (HRP)-conjugated goat anti-rabbit immunoglobulin (IgG) (catalog no. A6154), and HRP-conjugated goat anti-mouse IgG (catalog no. A4416) antibodies were obtained from Sigma-Aldrich. CQ (catalog no. C6628) was obtained from Sigma-Aldrich and dissolved in water. 3-MA was obtained from Tokyo Chemical Industry (catalog no. M2518) and was prepared fresh by dissolving in warm water (37 to 50°C) prior to treatment of cells. Rapamycin (LC Laboratories; catalog no. R-5000), Torin1 (Apex Bio; catalog no. A8312-5), and MRT68921 (Selleck Chemical; catalog no. S7949) were dissolved in dimethyl sulfoxide.

siRNA-mediated protein depletion.

mTOR, Rictor, and Raptor depletion was performed using small interfering RNA (siRNA) pools targeting mTOR (Dharmacon; catalog no. J-003008-11 and J-003008-12), Rictor (catalog no. J-016984-05 and J-016984-06), or Raptor (catalog no. J-004107-05 and J-004107-06). Cells were transfected with siRNAs at a final concentration of 80 nM using Lipofectamine RNAiMax (Invitrogen; catalog no. 13778030) following the manufacturer’s recommendation. A nontargeting (NT) siRNA (Qiagen; catalog no. 1027281) was used as a control. At 24 h posttransfection (hpt), the cells were replenished with DMEM containing 2% FBS and 1× antibiotics and were incubated for 48 h more prior to virus infection.

Cell lysate preparation and Western blotting.

Cells were washed with cold phosphate-buffered saline (PBS), treated with trypsin (Gibco; catalog no. 25200056), and pelleted at 500 × g for 5 min at 4°C. The cell pellet was resuspended in lysis buffer (25 mM Tris-HCl [pH 7.6], 150 mM NaCl, 1% Triton X-100, 1% sodium deoxycholate, 0.1% SDS) supplemented with the Halt protease and phosphatase inhibitor cocktail (Thermo Fisher; catalog no. 1861281) and incubated on ice for 5 min, after which the lysates were clarified by centrifugation at 10,000 × g for 5 min. Total protein quantification was done with a bicinchoninic acid (BCA) assay kit (Pierce; catalog no. 23225). Twenty-five micrograms of total protein was separated by electrophoresis on 6% or 8% polyacrylamide gels containing sodium dodecyl sulfate (SDS-PAGE). The separated proteins were transferred onto polyvinylidene difluoride (PVDF) membranes (Thermo Fisher; catalog no. 88520) using a Bio-Rad semidry transfer system. Nonspecific protein binding was blocked with 5% nonfat milk or with bovine serum albumin (for phosphorylated proteins) in Tris-buffered saline containing 0.2% Tween 20 (TBS-T) and incubated with the appropriate primary antibodies (dilutions, 1:5,000 for β-actin and 1:1,000 for all others) at 4°C overnight. Membranes were washed with TBS-T and incubated with appropriate HRP-conjugated secondary antibodies (dilution, 1:5,000) at room temperature (RT) for 2 h. Bands were detected using enhanced chemiluminescent (ECL) Western blotting substrate (Pierce; catalog no. 32106) in a Bio-Rad imager. Protein bands were analyzed with ImageJ software (NIH).

Determination of cell viability by flow cytometry.

Cell viability was determined by simultaneous determination of propidium iodide (PI; 1 μg/ml; Sigma-Aldrich; no. P4170) uptake, as a marker of plasma membrane integrity loss, and changes in the forward scatter properties of the cell (cell size) using flow cytometry (fluorescence-activated cell sorting). Flow cytometry and data analysis were performed as described in reference 42.

Confocal microscopy.

SK-N-SH cells stably expressing EGFP-LC3 (42) were mock infected or infected with ZIKV at an MOI of 1 in the presence or absence of CQ added 4 h prior to termination of the experiment. To visualize lysosomes, cells were incubated with 1 μM LysoTracker Red DND-99 (Invitrogen; catalog no. L7528) for 10 min, washed with PBS, and immediately imaged as described elsewhere (42). Fluorescent imaging was done in a Nikon A1R-Ti2 confocal system. Images were acquired using NIS-Elements software (Nikon). EGFP-LC3 puncta in cells were counted from three representative fields of images and were analyzed statistically.

Statistical analysis.

Experimental replicas were independent and performed on separate days. The data were analyzed by using one-way or two-way analysis of variance (ANOVA) and the appropriate post hoc test using the GraphPad Prism package. When ANOVA assumptions were not met (normality [Shapiro-Wilk test] or equal variance [Kruskal-Wallis]), one-way ANOVA on ranks or data transformation (two-way ANOVA) was performed on the data collected. Data were plotted as means and standard errors of the means (SEM) using the same package as for statistical analysis.