Abstract
Withdrawal of the growth factor interleukin-3 (IL-3) from IL-3-dependent myeloid cells causes them to undergo Bax/Bak1-dependent apoptosis, whereas factor-deprived Bax−/−Bak1−/− cells remain viable, but arrest and shrink. It was reported that withdrawal of IL-3 from Bax−/−Bak1−/− cells caused decreased expression of the glucose transporter Glut1, leading to reduced glucose uptake, so that arrested cells required Atg5-dependent autophagy for long-term survival. In other cell types, a decrease in Glut1 is mediated by the thioredoxin-interacting protein (Txnip), which is induced in IL-3-dependent myeloid cells when growth factor is removed. We mutated Atg5 and Txnip by CRISPR/Cas9 and found that Atg5-dependent autophagy was not necessary for the long-term viability of cycling or arrested Bax−/−Bak1−/− cells, and that Txnip was not required for the decrease in Glut1 expression in response to IL-3 withdrawal. Surprisingly, Atg5-deficient Bax/Bak1 double mutant cells survived for several weeks in medium supplemented with 10% fetal bovine serum (FBS), without high concentrations of added glucose or glutamine. When serum was withdrawn, the provision of an equivalent amount of glucose present in 10% FBS (~0.5 mM) was sufficient to support cell survival for more than a week, in the presence or absence of IL-3. Thus, Bax−/−Bak1−/− myeloid cells deprived of growth factor consume extracellular glucose to maintain long-term viability, without a requirement for Atg5-dependent autophagy.
Similar content being viewed by others
Introduction
Mammalian cells import a variety of extracellular nutrients to support their survival, growth, and proliferation. In vitro, cells are typically propagated in medium supplemented with glucose, glutamine, and fetal bovine serum (FBS). To generate chemical energy, glucose can be metabolized by oxidative phosphorylation in the mitochondria or by anaerobic glycolysis. Glutamine can be used as an energy source through glutaminolysis, and can also be used, together with a multitude of factors present in serum, to allow the synthesis of essential macromolecules for cell division.
Factor-dependent myeloid (FDM) progenitor cell lines [1, 2] require the hematopoietic growth factor interleukin-3 (IL-3) to survive and multiply in culture. When IL-3 is not provided, FDM cells undergo Bax/Bak1-dependent apoptosis. FDM cells derived from Bax−/−Bak1−/− mice survive in the absence of IL-3, but they undergo cell cycle arrest, similar to those over-expressing Bcl-2 [3, 4]. It has been reported that to survive longer than 4 days, IL-3-deprived Bax−/−Bak1−/− cells required Atg5-dependent autophagy [4], an intracellular catabolic process that can support the survival of starved or arrested cells [5]. However, given that our Atg5-depleted Bax/Bak1-null FDM cells survived for more than 8 days in the absence of IL-3 [6], it remains controversial whether growth factor-starved Bax−/−Bak1−/− FDM cells require Atg5-dependent autophagy for long-term survival.
Cells import glucose when it is available and growth factors can stimulate the amount of glucose that is consumed. Bax−/−Bak1−/− FDM cells arrested by IL-3 withdrawal decrease expression of the glucose transporter Glut1, which was proposed to reduce glucose uptake and metabolism [4]. Several mechanisms have been put forward to explain the regulation of expression and plasma membrane translocation of Glut1. Activation of autophagy was shown to be required for the efficient surface expression of Glut1 in mouse embryonic fibroblasts cultured under hypoxic conditions [6]. In human hepatocytes subjected to glucose starvation, a mechanism involving proteasomal degradation of the thioredoxin-interacting protein (Txnip) was required for Glut1 stabilization [7]. Moreover, it was shown that Txnip directly binds to Glut1 to promote its internalization, thus reducing Glut1 on the plasma membrane and the ability of cells to take up glucose [7].
In this study, we employed Bax−/−Bak1−/− FDM cells to genetically confirm that Atg5-dependent autophagy is not required for the sustained survival of arrested cells deprived of growth factor. We also asked whether autophagy and/or Txnip are required for the reduction in Glut1 expression and glucose uptake by arrested myeloid cells.
Results
Bax −/− Bak1 −/− myeloid cells shrink and arrest when deprived of IL-3, but survive
FDM cells require IL-3 to survive and proliferate [2, 4, 8]. When deprived of IL-3 for 4 days, >80% of wild-type Bax+/+Bak1+/+ FDM cells died, as indicated by loss of plasma membrane integrity and uptake of propidium iodide (PI) (Fig. 1a and Supplementary S1A). By contrast, and consistent with the findings of Lum et al. [4], Bax−/−Bak1−/− FDM cells survived in the absence of IL-3 for more than 14 days. At 48 h following IL-3 withdrawal, these cells shrank in size and arrested (Fig. 1b, c). This quiescent state was reversible, because when IL-3 was returned, the cells increased in size and resumed cycling. Thus, in the absence of Bax/Bak1-mediated apoptosis, FDM cells can survive for weeks without IL-3, but they shrink and arrest.
Bax −/− Bak1 −/− myeloid cells arrested by deprivation of IL-3 do not require Atg5-dependent autophagy to survive
During macroautophagy, LC3B-I is lipidated by a multiprotein complex containing Atg5 to form LC3B-II, which is incorporated into the growing autophagosome, and is commonly used as a marker of autophagy [9]. Lysates of Bax−/−Bak1−/− FDM cells, whether they were cultured in the presence or absence of IL-3, contained similar amounts of LC3B-II, suggesting that autophagy is occurring in both cycling and arrested cells (Fig. 2a). Notably, the low levels of both LC3B isoforms I and II in Bax−/−Bak1−/− cells cultured in serum, amino acid, and glucose-free Hank’s balanced salt solution (HBSS) is indicative of a strong increase in autophagic flux, which causes the coincident degradation of LC3B proteins after autophagosomes fuse with lysosomes [9].
To determine if Atg5-dependent autophagy was needed to support the sustained survival of IL-3-starved cells, we used CRISPR/Cas9 to delete Atg5 in Bax−/−Bak1−/− cells, generating Bax−/−Bak1−/−∆Atg5 triple-knockout FDM cell clones. Western blotting confirmed in three Bax−/−Bak1−/−∆Atg5 lines the absence of the LC3B-II isoform, whether they were cultured with or without IL-3, or in HBSS (Fig. 2a). Moreover, levels of LC3B-I in Bax−/−Bak1−/−∆Atg5 clones remained similar across all culture conditions. These results indicate that Atg5-dependent autophagy and autophagic flux are blocked in the Bax−/−Bak1−/−∆Atg5 lines. After 14 days of culture without IL-3, all Bax−/−Bak1−/−∆Atg5 clones remained ~80% viable (Fig. 2b). Similar to Bax−/−Bak1−/− FDMs, Bax−/−Bak1−/−∆Atg5 cells decreased in size and arrested within 48 h of growth factor withdrawal (Fig. 2c, d).
These results confirm that IL-3 receptor signals are needed for Bax−/−Bak1−/− FDM cells to maintain their normal size and divide, but show that Atg5-dependent autophagy is not required for their survival, whether they are cycling or are arrested by IL-3 withdrawal, consistent with our previous results using small interfering RNA (siRNA) to knockdown Atg5 in Bax−/−Bak1−/− FDM cells [10]. Furthermore, because culturing without IL-3 caused a similar decrease in size of the arrested Bax−/−Bak1−/− cells to that of the Bax−/−Bak1−/−∆Atg5 cells, this shrinkage is not due to Atg5-dependent autophagy.
Bax −/− Bak1 −/− myeloid cells arrested by IL-3 deprivation downregulate Glut1 independently of Txnip or Atg5 and survive in the absence of added glucose
Lum et al. [4] reported that when IL-3 growth factor was removed, Bax−/−Bak1−/− FDM cells decreased expression of the glucose transporter Glut1 [4]. A possible mechanism for this was proposed by Wu et al. [7], who found that Txnip could bind to Glut1 and induce its internalization through clathrin-coated pits, thus removing it from the plasma membrane [7]. Consistent with Lum et al. [4] and Wu et al. [7], levels of Txnip increased, and those of Glut1 decreased, in lysates of Bax−/−Bak1−/− cells 48 h following the removal of IL-3 (Fig. 3a).
To determine whether expression of Txnip caused or correlated with changes in Glut1 levels, we deleted Txnip in Bax−/−Bak1−/− FDM cells using CRISPR/Cas9. Western blotting confirmed the absence of Txnip protein in three Bax−/−Bak1−/−∆Txnip clonal lines (Fig. 3a). We found that Glut1 levels were unaffected by the deletion of Txnip, whether IL-3 was added or not, indicating that the increase in Txnip was not required for the decrease in Glut1 when cells were cultured without IL-3. Furthermore, similar to the Bax−/−Bak1−/− parental cells, withdrawal of IL-3 caused G1 arrest in Bax−/−Bak1−/−∆Txnip cells within 48 h (Supplementary Fig. S1A). Because induced overexpression of FLAG-Txnip in Bax−/−Bak1−/− cells had no impact on levels of Glut1 protein or cell cycle progression (Fig. 3b and Supplementary S1B), an increase in Txnip is not sufficient to reduce Glut1. Therefore, in FDM cells, the abundance of Txnip has no effect on levels of Glut1, whether they are treated with IL-3 or not. Of note, IL-3 withdrawal caused specific downregulation of Glut1, but not Glut4, in Bax−/−Bak1−/− and Bax−/−Bak1−/−∆Atg5 clones; and levels of Glut1 were similar among proliferating Bax/Bak1-null cells of all genotypes (Fig. 3a, b and Supplementary S1C). These results indicate that although IL-3 stimulates the expression of Glut1, neither Txnip nor Atg5-dependent autophagy is required for the decrease in Glut1 when growth factor is removed from FDM cells, and neither the absence nor overexpression of Txnip affects the levels of Glut1.
Because Glut1 levels are reduced in IL-3-starved Bax−/−Bak1−/− FDM cells, we asked whether glucose utilization was reduced. Proliferating Bax−/−Bak1−/− cells in the presence of IL-3 responded strongly to the addition of glucose, as indicated by the increased extracellular acidification rate (ECAR, Fig. 3c), showing that they readily consumed and metabolized the added glucose. Following 3 days of culture without IL-3, the rate of glycolysis in Bax−/−Bak1−/− cells was reduced, but they nevertheless still responded to the addition of glucose. With or without IL-3, the rates of glucose metabolism in Bax−/−Bak1−/− lines that also lacked Txnip or Atg5 were similar to the rates of the parental Bax/Bak1-null cells (Fig. 3c and Supplementary S1D). When cells were treated with oligomycin to block oxidative phosphorylation shortly after they reached their maximal glycolytic rate, the steady decline in ECAR continued relatively unchanged until glycolysis was inhibited with 2-deoxyglucose (2-DG), causing ECAR values to fall rapidly to baseline. These results indicate that FDM cells rely predominantly on glycolysis for energy production, and when deprived of IL-3, they retain the capacity to import and metabolize glucose, albeit at a reduced level.
Because IL-3-starved Bax−/−Bak1−/− FDMs took up and metabolized glucose, we hypothesized that this was sufficient to support their sustained survival despite the decrease in Glut1. FDM cells were routinely cultured in Dulbecco’s modified Eagle’s medium (DMEM) supplemented with 4 mM glutamine, 5.5 mM glucose, and 10% fetal bovine serum (FBS). The 10% FBS is estimated to also contain ~0.5 mM glucose (Supplementary Fig. S1E). To test their requirement for glucose, FDM cells were cultured in DMEM supplemented with 4 mM glutamine and 10% FBS, but lacking the added 5.5 mM glucose. Surprisingly, >80% of IL-3-starved Bax−/−Bak1−/− and Bax−/−Bak1−/−∆Txnip cells remained viable for 3 weeks in culture without added glucose (Fig. 3d). Moreover, when IL-3 was present, these cells were able to proliferate (Supplementary Fig. S1F). These results suggest that Bax−/−Bak1−/− FDM cells do not require high concentrations of glucose to maintain long-term viability, whether IL-3 is present or not, and that they consume the glucose in FBS and the added glutamine to survive or to proliferate with IL-3.
Bax −/− Bak1 −/− myeloid cells require glutamine to proliferate, but can survive in its absence for weeks
When glucose is limiting, lymphoid cells metabolize glutamine to survive and proliferate [11]. We therefore asked whether IL-3-dependent myeloid cells exhibited similar requirements for glutamine. In medium supplemented with 10% FBS but without added glucose or glutamine, with or without IL-3, >95% of wild-type Bax+/+Bak1+/+ FDM cells died as indicated by uptake of PI by day 4 (Supplementary Fig. S2A), and consistent with their death being due to apoptosis, the nuclei of the cells had sub-2n DNA (Fig. 4a). Apoptosis was Bax/Bak1-dependent, because Bax/Bak1-null cells, including clones also lacking Atg5, did not die when cultured in medium without added glucose and glutamine, but they arrested instead, even when IL-3 was provided (Fig. 4a). When cultured with neither added glucose nor glutamine, in the presence or absence of IL-3, >80% of the arrested Bax−/−Bak1−/− and Bax−/−Bak1−/−∆Atg5 cells survived for more than 12 days (Fig. 4b and Supplementary S2B).
When Bax−/−Bak1−/− cells were starved of added glucose and glutamine, most arrested in G1 with 2n DNA, but some arrested in G2/M with 4n DNA (Fig. 4a). When glutamine was added back to these arrested cells after 7 days, they resumed cycling within 48 h in the presence of IL-3, whereas adding back glucose had no effect (Supplementary Fig. S2C). These results indicate that IL-3-stimulated Bax−/−Bak1−/− myeloid cells require glutamine to proliferate, but can survive for weeks in its absence. Because Bax−/−Bak1−/−∆Atg5 cells behaved similarly to Bax−/−Bak1−/− parental cells, Atg5-dependent autophagy is not required for the sustained survival of IL-3-deprived Bax−/−Bak1−/− FDMs, even in cultures without high levels of added glucose and glutamine.
Bax −/− Bak1 −/− myeloid cells arrested by glutamine deprivation shrink in a manner independent of Atg5-mediated autophagy
Because Bax−/−Bak1−/− and Bax−/−Bak1−/−∆Atg5 cells arrested by IL-3 withdrawal decreased in size (Figs. 1b, 2c), we asked whether cells arrested by glutamine withdrawal also shrink. In medium supplemented with 10% FBS, but lacking added glucose and glutamine, IL-3-treated Bax−/−Bak1−/− and Bax−/−Bak1−/−∆Atg5 cells showed a gradual decrease in size to almost 1/3 of their original volume over a 20-day period (Fig. 5). Without IL-3, the shrinkage was more acute and cells reached ~1/3 of their original volume within 7 days. These results suggest that IL-3 provided a signal to Bax−/−Bak1−/− myeloid cells arrested by glutamine deprivation to maintain a large size, and confirm that the shrinkage of these cells upon IL-3 and/or nutrient withdrawal do not require Atg5-dependent autophagy.
Withdrawal of IL-3 or glutamine caused reduced expression of Mcl-1 and key regulators of cell cycle progression
To elucidate how growth factor and nutrient deprivation impacted FDM cell survival and proliferation, we performed Western blotting to determine protein levels of apoptotic and cell cycle regulators. We found that levels of the pro-survival protein Mcl-1 were markedly reduced in Bax−/−Bak1−/− and Bax−/−Bak1−/−∆Atg5 cells after 3 days in culture without added glucose and glutamine (Fig. 6a), suggesting that the decrease in normal levels of Mcl-1 allowed activation of Bax/Bak1 and the apoptosis of wild-type FDM cells observed in Fig. 4a. Consistent with this notion, addition of a specific Mcl-1 inhibitor S63845 [12] to Bax+/+Bak1+/+ cells was sufficient to cause significant apoptosis within 24 h, even when the cells were cultured with IL-3 in medium supplemented with glucose, glutamine, and 10% FBS (Fig. 6b).
As well as causing arrest, IL-3 withdrawal reduced c-Myc protein to barely detectable levels in both Bax−/−Bak1−/− and Bax−/−Bak1−/−∆Atg5 cells (Fig. 6a). Cyclin-dependent kinase 4 (Cdk4) and cyclin D3 are key drivers of cell cycle progression in hematopoietic cell lines [13, 14], and these proteins were also markedly reduced in the IL-3-starved, Bax/Bak1-deficient myeloid cells, suggesting a mechanism for their arrest (Figs. 1c, 2d). Withdrawal of glutamine from these cells also caused cell cycle arrest, even in the presence of IL-3 (Fig. 4a). Under these conditions, IL-3 was able to stimulate the expression of c-Myc and Cdk4, but the levels of these proteins were significantly lower than in the cells cultured with both glutamine and IL-3 (Fig. 6a). The transcription factor Atf4 accumulates in response to a variety of cellular stresses including amino acid starvation [15]. In glutamine-starved Bax−/−Bak1−/− and Bax−/−Bak1−/−∆Atg5 cells, Atf4 levels increase strongly, but intriguingly only in the presence of IL-3. These results indicate that the signal transduction pathway activated by ligation of the IL-3 receptors remains largely intact in cells cultured without supplemented glutamine, and can stimulate the cells to remain large, but that glutamine is essential for high levels of c-Myc, Cdk4, and cyclin D3 expression, and for the cells to divide.
Bax −/− Bak1 −/− myeloid cells consume extracellular glucose to maintain long-term viability
The Bax/Bak1-null cells cultured without high concentrations of added glucose and glutamine would need a source of energy to survive (Fig. 4b and Supplementary S2B), so we tested whether they were obtaining it from the supplemented FBS. In IL-3-containing medium lacking added glucose, glutamine, and FBS, all Bax−/−Bak1−/− and Bax−/−Bak1−/−∆Atg5 cells died after 2 days (Fig. 7a). In the absence of serum, we asked if adding back glutamine or glucose alone was sufficient to support cell survival. While <40% of cells cultured in medium supplemented with 4 mM glutamine remained viable after 2 days (and all died within 4 days, Supplementary Fig. S3A), >90% of cells cultured with 5.5 mM glucose survived for more than 6 days (Fig. 7a and Supplementary S3B). In IL-3-starved cultures, we observed a similar dependence on glucose, but not glutamine, for cell viability (Fig. 7b). Strikingly, 0.5 mM glucose only, an equivalent amount present in 10% FBS, was sufficient to support the survival of >80% of IL-3-deprived Bax−/−Bak1−/− and Bax−/−Bak1−/−∆Atg5 cells (Fig. 7b) for at least 1 week (Fig. 7c). Collectively, these results suggest that in the absence of growth factor, Bax−/−Bak1−/− myeloid cells consume extracellular glucose to maintain long-term viability, without the need to invoke Atg5-dependent autophagy.
Discussion
We have genetically eliminated Atg5 expression in IL-3-dependent Bax−/−Bak1−/− myeloid (FDM) cells and tested their ability to maintain survival in the absence of growth factor. Unlike Lum et al. [4], who concluded that Atg5-dependent autophagy was necessary to sustain the survival of IL-3-deprived Bax−/−Bak1−/− FDM cells beyond 4 days, our results demonstrate that canonical autophagy pathways are not required to support the long-term viability of these cells (Figs. 2, 4, 5, 7) [10]. We generated IL-3-dependent FDM lines by co-culturing E14.5 fetal liver single-cell suspensions from Bax−/−Bak1−/− embryos with fibroblasts expressing a HoxB8 retrovirus (to immortalize the cells) in the presence of IL-3 [16]. Lum et al. [4] obtained IL-3-dependent myeloid progenitors from the bone marrow of adult Bax−/−Bak1−/− mice, but did not specify how these cells were immortalized. It is plausible that differences in the age of donors, organ of origin, or immortalization protocols used to generate IL-3-dependent myeloid lines account for the different requirements for autophagy between our lines and those used by Lum et al. [4] when IL-3 was withdrawn. Furthermore, cell lines in culture can accumulate genetic and epigenetic changes [17], so we cannot exclude the possibility that such changes have altered the requirement of these lines for autophagy to survive when arrested. We also cannot exclude the possibility that the siRNAs used by Lum et al. [4] affected cell survival indirectly, for example, by induction of interferons. Nevertheless, because our IL-3-dependent Bax−/−Bak1−/− FDM lines did not require Atg5-dependent autophagy to survive when IL-3 was withdrawn, it is premature to generalize the conclusion of Lum et al. [4] that autophagy is essential for maintaining cell survival following growth factor withdrawal.
We were able to confirm the observation that culture of Bax−/−Bak1−/− FDM cells without IL-3 reduced the levels of the glucose transporter Glut1 [4], and wondered if the mechanism for reduction of Glut1 would be similar to that described by Wu et al. [7], who found that Txnip could reduce the levels of Glut1. Although we confirmed the reported inverse correlation between levels of Txnip and Glut1, Txnip is neither required nor sufficient for IL-3 to regulate levels of Glut1 in FDMs (Fig. 3). It remains possible that the Txnip-mediated downregulation of Glut1 is a mechanism specific to human hepatocytes and other cell types in response to glucose availability [7]. Canonical autophagy was shown to be required for the efficient surface expression of Glut1 [6]. Compared to parental Bax−/−Bak1−/− FDM cells, our Atg5-deficient Bax−/−Bak1−/− lines had similar levels of Glut1, decreased Glut1 to a similar extent when cultured without IL-3, and showed similar levels of glucose uptake to the parental cells when cultured with IL-3 or without. Therefore, in FDM cells, Atg5-dependent autophagy is not needed to maintain normal levels of Glut1, and does not affect the ability of IL-3-starved cells to take up and metabolize glucose.
Consistent with the requirement of glutamine for lymphoid cell survival [11], we found that wild-type FDM cells died following glutamine deprivation (Fig. 4). Surprisingly, glutamine was not required to support the survival of Bax/Bak1-null FDM cells in serum-containing cultures, and glutamine-deprived cells maintained a large size in the presence of IL-3, but arrested. This suggests that glutamine provided the source of amino acids needed for cell division, but that in its absence, Bax−/−Bak1−/− cells import and metabolized the glucose present in serum to maintain viability. We also found that the shrinkage of FDM cells arrested by glutamine and/or IL-3 withdrawal occurred independently of Atg5-mediated autophagy (Fig. 5), consistent with Hecht et al. [18], who reported that the acute shrinkage of B and T lymphocytes following growth factor withdrawal occurred before the induction of autophagy.
In the absence of high levels of glutamine and glucose, pathways activated by IL-3 could still converge to induce the expression of c-Myc in Bax/Bak1-null FDM cells, and although this response was blunted, the arrested cells maintained a relatively large size for at least 1 week (Fig. 5). The transcription factor Atf4 was strongly induced by IL-3 in glutamine-starved cells, but it is unclear how or why signaling from the IL-3 receptor leads to increased Atf4 production. Although Atf4 was shown to activate the expression of several autophagy genes to promote cell survival [19], we have not been able to confirm that its accumulation induced LC3B lipidation in the glutamine-starved FDM cells. That IL-3 also induced Atf4 expression in Atg5-deficient Bax−/−Bak1−/− cells following glutamine deprivation (Fig. 6) suggests that Atf4 does not promote canonical autophagy to maintain cell viability under these conditions. Moreover, cells starved of both IL-3 and glutamine showed no detectable expression of Atf4, yet survived for weeks.
In the absence of IL-3 and high concentrations of added glucose and glutamine, Atg5-deficient Bax−/−Bak1−/− cells remained viable for weeks in serum-containing medium, and their rapid death in glucose-free cultures was consistent with the hypothesis that they took up and used the glucose present in serum to survive (Fig. 7). That these cells predominantly generate chemical energy through glycolysis (Fig. 3) was also consistent with their ability to consume only diluted concentrations of glucose to remain viable for at least a week. Serum-starved Bax/Bak1-null cells supplemented with high concentrations of glucose and glutamine survived for several weeks, but arrested, with or without IL-3 (data not shown). Collectively, our work suggests that in the absence of growth factor, Bax−/−Bak1−/− FDM cells, including mutant lines lacking Atg5-dependent autophagy, import extracellular glucose and amino acids to maintain long-term viability. The metabolic requirement for glucose and amino acids is a mechanism to support cell survival, even in cultures where levels of these nutrients are not optimal for cell proliferation, and the reduced levels of c-Myc, Cdk4, and cyclin D3, combined with possibly altered levels of other cell cycle regulators, could explain why growth factor and/or nutrient-starved Bax−/−Bak1−/− FDM cells arrested (Fig. 8).
Materials and methods
Cell culture
Wild-type Bax+/+Bak1+/+ and Bax−/−Bak1−/− IL-3-dependent mouse myeloid progenitor (FDM) cell lines have been described [16]. We routinely test our cells for mycoplasma contamination. Cells were cultured in DMEM (Thermo Fisher Scientific, Cat# 11885-084; containing 5.55 mM d-glucose and 4 mM l-glutamine) supplemented with 10% FBS (Sigma, Cat# F9423) and 0.25 ng/mL recombinant murine IL-3 (PeproTech, Cat# 213-13) in a 37 °C humidified incubator with 10% CO2. Cells were deprived of IL-3, glucose, glutamine, glucose and glutamine, or varying concentrations of FBS, by washing three times with medium that lacked those nutrients, and then resuspended in the same medium for analysis at the indicated time points. Unless indicated otherwise, 1 × 105 cells were seeded in 1 mL of the specified medium per well of a 24-well tissue culture plate (Corning, Cat# 353047), and one well was analyzed for each timepoint. Media were not renewed or topped up during time-course experiments. To prevent cells from sticking to the plasticware in experiments without FBS, tissue culture dishes were pre-coated with 20 μg/mL bovine serum albumin (Sigma, Cat# A1470) and rinsed with serum-free medium before cell seeding. DMEM lacking d-glucose but containing 4 mM l-glutamine (Cat# 11966-025) and DMEM lacking both d-glucose and l-glutamine (Cat# A14430-01) were purchased from Thermo Fisher Scientific. d-Glucose was purchased from Sigma (Cat# G7528). The glucose concentration in FBS was determined using an Accu-Chek Performa glucose meter, according to the manufacturer’s instructions.
Lentiviral constructs for CRISPR/Cas9 and inducible overexpression
An inducible lentiviral CRISPR/Cas9 system [20] was used to mutate Txnip in Bax−/−Bak1−/− FDM cells. Guide RNA oligos targeting mouse Txnip (guide 1: CCA GCA GGT GAG AAC GAG A; guide 2: TAT GGT TGC GTA GAC TAC T) were cloned into the BsmBI site of the pFgH1tUTG GFP lentiviral vector. Guide RNA oligo targeting mouse Atg5 (AAG ATG TGC TTC GAG ATG TG) was cloned into the BsmBI site of the lentiviral vector lentiCRISPR v2 (Addgene, #52961 [21]). For overexpression studies, the coding sequence of mouse Txnip was PCR amplified from pEGFP-C1-Txnip (Addgene, #18758 [22]) with the incorporation of an N-terminal FLAG tag. The PCR product was cloned into the EcoRI/NheI sites of a doxycycline-inducible pFTREtight MCS rtTAadvanced GFP lentiviral vector [23].
Antibodies and small molecules
Antibodies: LC3B (D11, Cell Signaling Technology, #3868), Atg5 (D5F5U, Cell Signaling Technology, #12994); β-actin (AC-15, Sigma, #A1978), Glut1 (Abcam, ab15730), Txnip (D5F3E, Cell Signaling Technology, #14715 S), Glut4 (1F8, Cell Signaling Technology, #2213), Mcl-1 (Rockland, 600-401-394S), c-Myc (Abcam, ab32072), Cdk4 (Santa Cruz Biotechnology, sc-260), Cyclin D3 (Cell Signaling Technology, #2936), and Atf4 (D4B8, Cell Signaling Technology, #11815). Small molecules: Mcl-1 inhibitor S63845 was synthesized according to Kotschy et al. [12] by SYNthesis Med Chem Pty Ltd (Parkville, Australia) or by Active Biochem (Catalog# A-6044—a kind gift from David Huang); dimethyl sulfoxide (Cat# 472301) and propidium iodide (Cat# P4864) were purchased from Sigma.
Measurements of cell death, cell cycle, and cell size
Cell death and cell cycle were assayed using propidium iodide (PI) and flow cytometry. For death assays, cells were harvested and resuspended in phosphate-buffered saline (PBS) containing 1 μg/mL PI. For cell cycle analysis, cells were harvested and resuspended in PBS containing 0.1% Triton X-100, 5 μg/mL PI, and 10 μg/mL RNaseA (Thermo Scientific), and then incubated at 37 °C for 15 min. Samples were processed on FACSCalibur or LSRFortessa X-20 flow cytometers (BD Biosciences) and the acquired data analyzed with the FlowJo software. Cell volume in Fig. 5 was measured using a CASY Coulter-based cell counter (Roche Innovatis AG). In brief, 200 μL aliquots of cells cultured under the specified conditions were mixed with 5 mL of CASY electrolyte buffer and analyzed in a 60 μm CASY capillary. Multiple parameters including the peak cell volume were collected.
Glucose uptake assays
ECAR rates were measured in live cells using a Seahorse Bioscience XFe96 Analyzer (Agilent Technologies). In brief, cells were resuspended in non-buffered DMEM (Seahorse Biosciences) and 200 × 103 cells were seeded per well in Seahorse Bioscience 96-well plates. The cellular ECAR was analyzed in non-buffered DMEM containing 2 mM glutamine with the following compounds: 5 mM glucose, 1 μM oligomycin, and 50 mM 2-deoxyglucose. For each assay cycle, measurement time points of 2 min mix, 2 min wait, and 5 min measure were collected.
Data reproducibility
Unless indicated otherwise, all experiments shown have been reproduced at least three independent times with similar results. While not all repeats were performed wholly or in exact detail, we ensured that our experimental design allowed us to interpret and draw conclusions from multiple independent experiments. This is aided by the fact that our results are unambiguous: the majority of cells are either alive or dead, cycling or arrested, large or small, and immunoblots show clear changes in protein levels or modification. We have endeavored to perform experiments that provide consistent evidence in mechanistically independent ways, rather than repeating the same experiment many times until statistical significance is reached. The experimental triangulation of our evidence [24] is best exemplified by multiple independent assays confirming that autophagy-deficient FDM cells survived for weeks without growth factor under a variety of different culture conditions (Figs. 2b, 4b, 5, 7c and Supplementary Fig. S3B).
References
Perkins AC, Cory S. Conditional immortalization of mouse myelomonocytic, megakaryocytic and mast cell progenitors by the Hox-2.4 homeobox gene. EMBO J. 1993;12:3835–46.
Ekert PG, Read SH, Silke J, Marsden VS, Kaufmann H, Hawkins CJ, et al. Apaf-1 and caspase-9 accelerate apoptosis, but do not determine whether factor-deprived or drug-treated cells die. J Cell Biol. 2004;165:835–42.
Vaux DL, Cory S, Adams JM. Bcl-2 gene promotes haemopoietic cell survival and cooperates with c-myc to immortalize pre-B cells. Nature. 1988;335:440–2.
Lum JJ, Bauer DE, Kong M, Harris MH, Li C, Lindsten T, et al. Growth factor regulation of autophagy and cell survival in the absence of apoptosis. Cell. 2005;120:237–48.
Mizushima N, Komatsu M. Autophagy: renovation of cells and tissues. Cell. 2011;147:728–41.https://doi.org/710.1016/j.cell.2011.1010.1026.
Roy S, Leidal AM, Ye J, Ronen SM, Debnath J. Autophagy-dependent shuttling of TBC1D5 controls plasma membrane translocation of GLUT1 and glucose uptake. Mol Cell. 2017;6:020.
Wu N, Zheng B, Shaywitz A, Dagon Y, Tower C, Bellinger G. et al. AMPK-dependent degradation of TXNIP upon energy stress leads to enhanced glucose uptake via GLUT1. Mol Cell. 2013;49:1167–75.https://doi.org/1110.1016/j.molcel.2013.1101.1035.
Dexter TM, Garland J, Scott D, Scolnick E, Metcalf D. Growth factor-dependent hemopoietic precursor cell lines. J Exp Med. 1980;152:1036–47.
Loos B, du Toit A, Hofmeyr JH. Defining and measuring autophagosome flux-concept and reality. Autophagy. 2014;10:2087–96.
Lindqvist LM, Heinlein M, Huang DC, Vaux DL. Prosurvival Bcl-2 family members affect autophagy only indirectly, by inhibiting Bax and Bak. Proc Natl Acad Sci USA. 2014;111:8512–7.https://doi.org/8510.1073/pnas.1406425111.
Le A, Lane AN, Hamaker M, Bose S, Gouw A, Barbi J. et al. Glucose-independent glutamine metabolism via TCA cycling for proliferation and survival in B cells. Cell Metab. 2012;15:110–21.https://doi.org/110.1016/j.cmet.2011.1012.1009.
Kotschy A, Szlavik Z, Murray J, Davidson J, Maragno AL, Le Toumelin-Braizat G, et al. The MCL1 inhibitor S63845 is tolerable and effective in diverse cancer models. Nature. 2016;538:477–82.
Dolezal E, Infantino S, Drepper F, Borsig T, Singh A, Wossning T, et al. The BTG2-PRMT1 module limits pre-B cell expansion by regulating the CDK4-Cyclin-D3 complex. Nat Immunol. 2017;18:911–20.
Sawai CM, Freund J, Oh P, Ndiaye-Lobry D, Bretz JC, Strikoudis A, et al. Therapeutic targeting of the cyclin D3:CDK4/6 complex in T cell leukemia. Cancer Cell. 2012;22:452–65.
Harding HP, Zhang Y, Zeng H, Novoa I, Lu PD, Calfon M, et al. An integrated stress response regulates amino acid metabolism and resistance to oxidative stress. Mol Cell. 2003;11:619–33.
Ekert PG, Jabbour AM, Manoharan A, Heraud JE, Yu J, Pakusch M, et al. Cell death provoked by loss of interleukin-3 signaling is independent of Bad, Bim, and PI3 kinase, but depends in part on Puma. Blood. 2006;108:1461–8.
Ben-David U, Siranosian B, Ha G, Tang H, Oren Y, Hinohara K, et al. Genetic and transcriptional evolution alters cancer cell line drug response. Nature. 2018;560:325–30.
Hecht VC, Sullivan LB, Kimmerling RJ, Kim DH, Hosios AM, Stockslager MA, et al. Biophysical changes reduce energetic demand in growth factor-deprived lymphocytes. J Cell Biol. 2016;212:439–47.
B’Chir W, Maurin AC, Carraro V, Averous J, Jousse C, Muranishi Y, et al. The eIF2alpha/ATF4 pathway is essential for stress-induced autophagy gene expression. Nucleic Acids Res. 2013;41:7683–99.
Aubrey BJ, Kelly GL, Kueh AJ, Brennan MS, O’Connor L, Milla L, et al. An inducible lentiviral guide RNA platform enables the identification of tumor-essential genes and tumor-promoting mutations in vivo. Cell Rep. 2015;10:1422–32.
Sanjana NE, Shalem O, Zhang F. Improved vectors and genome-wide libraries for CRISPR screening. Nat Methods. 2014;11:783–4.
Wang Z, Rong YP, Malone MH, Davis MC, Zhong F, Distelhorst CW. Thioredoxin-interacting protein (txnip) is a glucocorticoid-regulated primary response gene involved in mediating glucocorticoid-induced apoptosis. Oncogene. 2006;25:1903–13.
Brumatti G, Salmanidis M, Kok CH, Bilardi RA, Sandow JJ, Silke N, et al. HoxA9 regulated Bcl-2 expression mediates survival of myeloid progenitors and the severity of HoxA9-dependent leukemia. Oncotarget. 2013;4:1933–47.
Munafo MR, Davey Smith G. Robust research needs many lines of evidence. Nature. 2018;553:399–401.
Acknowledgements
We thank Gabriela Brumatti, Rebecca Feltham, James Vince, Simon Monard, David Huang, and Andre Samson for discussion, technical advice, and reagents. Funding for this project was provided by the Australian National Health and Medical Research Council (NHMRC) Program Grant #1113133 and NHMRC fellowship 1020136 to DLV; and the Leukemia and Lymphoma Society SCOR grant #7001-13. Work in the authors’ laboratory is made possible by operational infrastructure grants through the Australian Government Independent Research Institutes Infrastructure Support (IRISS) and the Victorian State Government OIS.
Author information
Authors and Affiliations
Corresponding author
Ethics declarations
Conflict of interest
The authors declare that they have no conflict of interest.
Additional information
Publisher’s note: Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
Edited by E. Baehrecke
Supplementary information
Rights and permissions
About this article
Cite this article
Dong, L., Reljic, B., Cheung, J.G. et al. In the absence of apoptosis, myeloid cells arrest when deprived of growth factor, but remain viable by consuming extracellular glucose. Cell Death Differ 26, 2074–2085 (2019). https://doi.org/10.1038/s41418-019-0275-z
Received:
Revised:
Accepted:
Published:
Issue Date:
DOI: https://doi.org/10.1038/s41418-019-0275-z