Recent work, especially in the yeast Saccharomyces cerevisiae, has demonstrated that mRNA movement from active translation to cytoplasmic granules, termed mRNA‘p-bodies’ (processing bodies), occurs in concert with the regulation of translation during cell stress. However, the signals regulating p-body formation are poorly defined. Recent results have demonstrated a function for sphingolipids in regulating translation during heat stress, which led to the current hypothesis that p-bodies may form during heat stress in a sphingolipid-dependent manner. In the present study, we demonstrate that mild-heat-stress-induced formation of p-bodies, as determined by localization of a GFP (green fluorescent protein)-tagged Dcp2p and RFP (red fluorescent protein)-tagged Edc3p to discrete cytoplasmic foci. Sphingoid base synthesis was required for this effect, as inhibition of sphingoid base synthesis attenuated formation of these foci during heat stress. Moreover, treatment of yeast with the exogenous sphingoid bases phyto- and dihydro-sphingosine promoted formation of p-bodies in the absence of heat stress, and the lcb4/lcb5 double-deletion yeast, which accumulates high intracellular levels of sphingoid bases, had large clearly defined p-bodies under non-stress conditions. Functionally, inhibition of sphingolipid synthesis during heat stress did not prevent translation stalling, but extended translation arrest, indicating that sphingolipids mediate translation initiation. These results are consistent with the notion that p-bodies serve not only in mRNA degradation, but also for re-routing transcripts back to active translation, and that sphingolipids play a role in this facet of the heat-stress response. Together, these results demonstrate a critical and novel role for sphingolipids in mediating p-body formation during heat stress.
- heat-stress response
- processing body
- Saccharomyces cerevisiae
A central component of the cellular response to stress is the induction of large-scale changes in specific mRNA transcript levels [1,2]. Although transcriptional changes in response to stress conditions are well established, an appreciation has recently developed for post-transcriptional mechanisms of regulating mRNA, such as mRNA sequestration to cytoplasmic granules including mRNA ‘p-bodies’ (processing bodies) [3,4]. In yeast, these cytoplasmic foci form during stress conditions including glucose depletion, salt treatment and late stages of culture growth [5,6]. P-bodies comprise mRNA, RNA-interacting proteins, including the core p-body components involved in cap-dependent degradation (Dcp1p, Dcp2p, Edc3p and others), and may contain proteins involved in translation and/or its regulation [7,8]. Although current models place translation arrest upstream from p-body formation (reviewed in ), several lines of evidence in yeast and mammalian cells indicate that mRNA from p-bodies can return to translation, thus placing p-bodies themselves upstream from translation initiation [3,9]. This has led to the concept of an ‘mRNA cycle’ whereby mRNA moves dynamically through translation, p-bodies and stress granules, although directionality has not been firmly established .
Recent decades have witnessed the emergence of sphingolipids, including sphingoid bases and ceramide, as key mediators of stress responses [11,12]. Biosynthesis of sphingolipids occurs in response to cytokines, UV irradiation, DNA-damaging chemicals and many other stressors including thermal stress [11,12]. Thus the HSR (heat-stress response) of Saccharomyces cerevisiae has been used as a model system for dissecting eukaryotic stress responses [13–15], including the role of sphingolipids in mediating HSR subprogrammes such as transcriptional reprogramming , translation of heat-shock proteins  and protein degradation [18,19]. These roles of sphingolipids in stress responses, along with studies implicating sphingolipid synthesis in translational regulation during heat stress, suggested the hypothesis that sphingolipids may be linked mechanistically and functionally to p-body formation in yeast. Although p-body formation in response to heat has not been previously reported to occur in S. cerevisiae, given heat stress transiently decreases translation initiation , it seemed possible that a concomitant increase in p-bodies may occur in response to heat stress.
In the present study, we show that heat stress induces p-body markers to discrete cytoplasmic foci. Mechanistically, the results demonstrate that this occurs in a sphingolipid-dependent manner. These results also provide insights into the roles of bioactive sphingolipids in mediating translational control during the HSR. Thus these results provide important insights into the mechanisms of p-body formation, as well as their functions, and they also underscore the fundamental and multiple roles played by members of this lipid class in cellular adaptation to stress.
Yeast strains, culture and treatment
Genotypes of yeast strains used in this study are listed in Table 1. The strain bearing a GFP (green fluorescent protein)-tagged eIF3 (eukaryotic initiation factor 3) protein expressed from its chromosomal location was purchased from Invitrogen. Routine culture of yeast was carried out in yeast YPD medium [1% (w/v) yeast extract/2% (w/v) proteose peptone/2% (w/v) dextrose] at 30 °C with 200–250 rev./min of shaking. Edc3p-mCherry was expressed from the pRP1574 vector  and selected on uracil drop-out medium. Unless otherwise noted, strains were grown to mid-exponential phase (D600=0.4–0.6) prior to treatment. Phytosphingosine, sphingosine and stearylamine were purchased from Sigma–Aldrich, and all other lipids were synthesized and purified at the Lipidomics Core Facility at the Medical University of South Carolina.
After treatment, cells were quickly pelleted in a bench-top microcentrifuge (13000 g) for 15–30 s. Medium was decanted, and cell pellets were resuspended in residual medium, and immediately spotted on to microscope slides. As yeast under coverslips are subject to hypoxic stress , coverslips were applied to droplets only immediately prior to visualization. Cells were visualized by confocal microscopy using an LSM510 microscope (Carl Zeiss). Images show a single plane of the sample.
Sphingoid base measurements
After treatment, cell pellets were harvested by centrifugation at 3500 g for 3 min. Pellets were immediately resuspended in 3 ml of chloroform/methanol at a ratio of 2:1, and sphingolipids were extracted as described previously . The lipid fraction was subjected to mild alkaline hydrolysis. Sphingoid bases were converted into fluorescent o-pthalaldehyde derivatives, and resolved by HPLC as described previously . Measurements were normalized to total phospholipid determined in an aliquot of the lipid extract as described previously .
Yeast cytoplasmic extracts were prepared from exponential-phase cultures treated as described previously . Cells were treated as indicated, and then incubated for 15 min at room temperature (23 °C) with 0.1 mg/ml cycloheximide added to the culture medium. Cells were collected by centrifugation (5000 g for 5 min) and suspended in 1 ml of lysis buffer (10 mM Tris/HCl, pH 7.5, containing 100 mM NaCl, 30 mM MgCl2, 0.1 mg/ml cylcoheximide and 0.2 mg/ml heparin). Cells were disrupted by vortexing with glass beads, and 20 attenuance units were loaded on to a 15–50% sucrose gradient buffered with 50 mM Tris/acetate (pH 7.4), 50 mM NH4Cl, 12 mM MgCl2 and 1 mM DTT (dithiothreitol). Gradients were centrifuged in an SW41 rotor at 40000 rev./min for 120 min at 4 °C. After ultracentrifugation, the tubes were punctured at the bottom, and the fractions were displaced upwards using Fluorinert FC-40 (Sigma–Aldrich). Absorbance was monitored continuously at 254 nm using ISCO-UA45 (model 640) density-gradient fraction collector. The area under the curve was calculated by printing profiles on to graph paper and counting the area under the free 80S peak and the first six polyribosome peaks.
Heat stress induces p-body formation
The protein comprising the mRNA decapping enzyme Dcp2p participates in mRNA degradation, and it localizes in p-bodies in response to several conditions including glucose deprivation and late growth stage [4,6]. Thus GFP-tagged Dcp2p has been used as a marker for p-bodies . In a strain containing a GFP tag sequence in the chromosomal DCP2 gene, to permit p-body visualization by confocal microscopy in live cells, p-bodies were undetectable during exponential-phase growth under normal conditions, but stress conditions including growth to saturation leads to concentrations of GFP signal in several punctuate cytoplasmic foci . Consistent with these results, approx. 1–3 cytoplasmic foci of approx. 1 μm in diameter were observed in yeast grown to saturation at 30 °C in YPD medium, but not in cells in growing exponentially (Figure 1A). In general, the intensity, duration and type of stress regulates p-body size and number .
Owing to the highly conserved nature of the HSR in S. cerevisiae, we sought to investigate whether heat stress would induce similar Dcp2p–GFP localization. Cells were grown in rich medium to mid-exponential phase (D600=0.3–0.6), and then aliquots of live cells were quickly mounted and visualized by confocal microscopy over a time course. Indeed, shifting exponential-phase cultures from the normal growth temperature of 30 °C to the mild heat-stress temperature of 39 °C caused formation of faint punctuate spots of a comparatively smaller size, approx. 0.5 μm, than those observed in saturated cultures (Figure 1B). These faint foci first appeared after 10 min of heat stress. Over 10–20 min these foci increased in apparent fluorescence intensity, and the number of cells displaying these foci increased, such that by 20 min, 60% of the cells showed many very faint foci, with a small number of cells (<10%) showing 1 or more foci of 0.5 μm in diameter. In cells held at 39 °C, dissolution of the signal occurred by 60 min (Figure 1B).
This yeast strain provided a distinct advantage over heterologous expression of p-body markers, in that this strain did not require growth on selection medium, which, in our hands, caused a small amount of p-body formation under conditions of logarithmic growth in some genetic backgrounds. On the other hand, it was important to employ another independent p-body marker. Thus we utilized Edc3p, which activates mRNA decapping and, along with enzymes required for decapping and mRNA degradation, serves as a core component of p-bodies. An Edc3p construct tagged with the RFP (red fluorescent protein) mCherry was expressed on a low-copy plasmid selectable in uracil drop-out medium. Yeast transformed with this construct showed very few p-bodies during exponential-phase growth at 30 °C. In contrast, a shift in culture temperature to 39 °C for 25 min precipitated a clear-cut induction of p-bodies (Figure 1C), which were similar in size and number to those observed using the Dcp2p–GFP-expressing strain (Figure 1B). Expression of the Edc3p construct in the Dcp2–GFP strain demonstrated that these two tagged proteins co-localized (results not shown). The observation of p-bodies using two different markers and two different background strains supported the hypothesis that heat stress induces p-body formation in yeast, albeit both greater in number and smaller in size than those observed in saturated yeast cultures.
A recent study indicated that ‘robust’ heat shock (46 °C) induced formation of cytoplasmic foci containing the translation initiation factor eIF3a/Tif32p/Rpg1p and other eIF subunits . In that study, the authors demonstrated that these granules contained similar protein components as stress granules, which were previously identified in mammalian cells. Thus the authors concluded that these eIF3-containing granules represent yeast stress granules. Intriguingly, these granules also contained p-body markers, including Dcp2p. Therefore, it became essential to determine whether the mild-heat-stress-induced foci observed in the present study bear any relation to the eIF3-containing stress granules. To test this, we obtained a yeast strain bearing GFP-tagged eIF3a, Tif32p and Rpg1p , now commercially available, and transfected it with either empty vector or the vector encoding the Edc3p–RFP construct and visualized these strains as above. Images indicated that yeast transformed with empty vector demonstrated an evenly distributed cytosolic localization of eIF3a, consistent with previous reports (Figure 1D) [8,23]. Furthermore, when these yeast were transfected with the Edc3p–RFP construct, a few red foci were clearly observed at control temperature of 30 °C (presumably due to the selection medium, which, in our hands, induced p-body formation in some genetic backgrounds). Shifting these cells to the mild heat-stress temperature of 39 °C induced formation of Edc3p–RFP foci in most cells; in contrast, the stress granule marker eIF3a remained largely cytosolic in an even distribution and without significant foci formation under the mild heat-stress employed, consistent with the previous study . Thus we concluded that these Dcp2p/Edc3p containing foci are distinct from yeast stress granules and represent bona fide p-bodies.
Heat-induced p-body formation requires de novo sphingolipid synthesis
We and others demonstrated previously an acute flux of de novo sphingolipid synthesis at early time points of heat stress [25,26]. This flux induces transient increases in sphingolipids over a time course of heat stress, through increasing serine supply to serine palmitoyltransferase . Initially, increases occur in the sphingoid bases dihydro- and phyto-sphingosine, with subsequent elevation in sphingoid base phosphates and ceramides [25,26,28]. Although this phenomenon is observed among a wide variety of yeast strains, we reported previously significant variation between parental wild-type strains in the magnitude of this aspect of the HSR . Thus we tested the Dcp2–GFP-expressing strain for its ability to increase sphingoid bases in response to heat stress. Indeed, this strain demonstrated a robust 2–5-fold increase in sphingoid base species by 10 min of heat stress, which was completely attenuated by a 5-min pre-treatment with the serine palmitoyltransferase inhibitor, myriocin (Figure 2A). Thus we used these conditions to test whether sphingoid base production during heat stress was a requirement for p-body formation. Pre-treatment with myriocin significantly blunted formation of p-bodies at 20 and 30 min following heat stress (Figure 2B). These results indicate that de novo sphingolipids are critical for p-body formation during heat stress.
Treatment with sphingoid bases induces p-body formation
Although heat stress initiates a cascade of sphingolipid synthesis, the sphingoid bases phyto- and dihydro-sphingosine are well established signalling sphingolipids in yeast and their levels increase in a time frame commensurate with the formation of p-bodies [25,26]. In an attempt to identify the active species for heat-mediated p-body formation, cells at the non-stress temperature of 30 °C were treated with a panel of sphingoid bases or similar compounds at 10 μM. Treatment of cells with yeast sphingoid base species initiated the localization of Dcp2–GFP to cytoplasmic foci, with phytosphingosine being the most active, followed by D-erythro-dihydrosphingosine, and then its unnatural steroisomer, L-threo-dihydrosphingosine (Figure 3). In contrast, sphingosine, the major sphingoid base in mammalian cells, was unable to induce p-body formation. Likewise, neither ceramide nor stearylamine caused p-bodies to form (Figure 3). Thus yeast sphingoid bases, and particularly phytosphingosine, seem the likely candidates for mediating p-body formation during heat stress.
Sphingoid bases undergo conversion into phosphorylated derivatives that may play roles in growth and mitochondrial function [29,30]. Moreover, phosphorylation of sphingoid bases serves to regulate their levels, as deletion of the sphingoid base kinases Lcb4p and Lcb5p, increases sphingoid base levels severalfold . Indeed, during exponential growth at 30 °C, cells deleted for these kinases demonstrated well over 10-fold increases in both C18 and C20 sphingoid base species (Figure 4A). Thus this strain provided a useful tool in which to test whether elevation of endogenous sphingoid bases is sufficient to induce p-bodies even under normal growth conditions, and, moreover, whether their phosphorylation is required for p-body formation. The lcb4Δ/lcb5Δ double-deletion strain and its parental wild-type strain JK93dα were both transformed with the Edc3–mCherry construct. Transformed strains were grown to a D600 0.7 in selective medium at 30 °C and visualized with confocal microscopy. Indeed, under these conditions, fluorescent foci were virtually undetectable in the parental background strain. In contrast, the lcb4Δ/lcb5Δ double-deletion strain showed 1–2 relatively large p-bodies in most cells (Figure 4B). These results further connect sphingolipid metabolism to p-body formation and indicate that elevating sphingoid bases increases p-bodies, even in the absence of heat stress. Additionally, these results confirm that further metabolism of sphingoid bases to their phosphorylated derivatives is not required for sphingolipid-mediated p-body formation.
Myriocin potentiates heat-mediated translation arrest
Current results imply that formation of p-bodies is downstream from a stress-induced stall in translation initiation [6,8]; however, previous data indicated a role of sphingoid bases in increasing translation initiation during heat stress, after the initial stall . Thus we sought to determine whether sphingoid bases mediate translation arrest or, as an alternative, whether inhibiting sphingolipid synthesis block p-body formation and subsequently delay the recovery of translation initiation.
In this effort, the Dcp2–GFP-expressing strain (yRP1727) was treated with myriocin or vehicle control at the onset of heat stress, and active translation was determined by analysis of polyribosome profiles. Equal protein (20 attenuance units) was loaded on to sucrose gradients and analysed as described in the Experimental section. The results (Figure 5) show that exponential-phase-growing yeast at 30 °C demonstrated significant RNA distribution in polysomes relative to the free 80S ribosomal subunit (73% compared with 27% total peak area respectively), whereas heat stress for 15 min caused a decrease in polysome peak area and a concomitant increase in the area under the free subunit peak (35% compared 65% total peak area respectively), reflecting a decrease in active translation. Treatment with myriocin during heat stress moderately enhanced this decrease, as demonstrated by the even greater area under the free ribosomal subunit peaks relative to the polysome peaks (30% compared with 70% total peak area respectively). These results agree with previous findings indicating a key role for sphingoid bases in regulation of translation initiation following stress  and thus suggest that in addition to sequestering mRNA upon translation arrest, they also mediate the later recovery of active translation. This hypothesis has been called the ‘mRNA cycle’ by Parker and colleagues, in such that stress-induced formation of cytoplasmic mRNP (messenger ribonucleoprotein) granules, such as p-bodies and the related stress granules, may store mRNA species during stress conditions thus facilitating the recovery of translation [3,9].
The present study provides results demonstrating that heat stress induces p-body formation in S. cerevisiae, and also implicates the acute induction of sphingoid base production in this process. The results also support the notion of a requirement for p-bodies in resumption of translation after heat stress, and suggest that sphingolipids may mediate translation initiation primarily through promoting p-body formation.
To our knowledge, these data represent the first observation of p-body formation during heat stress in yeast, although p-body formation in response to heat has been observed in mammalian cells . Although one report indicated that stress granules, which are thought to form subsequently to p-bodies, have been observed in yeast , and that these foci contained typical p-body markers, these foci were observed only under severe heat-shock conditions (46 °C for 10 min) and did not occur at lower temperatures. We show in the present study that p-bodies form during heat stress in yeast, although they were smaller and more faint than those formed under other stress conditions including stationary phase growth and/or glucose deprivation. Thus p-body formation is emerging as a generalized response to various stress stimuli. Although recent studies have demonstrated that p-bodies are required for subsequent formation of stress granules , the precise directionality of mRNA flow through polysomes and cytosolic mRNP particles remains unclear at present . Moreover, it is also unclear if mRNA dynamics differ on the basis of the specific stress stimulus and/or the species under study.
Sphingolipid biosynthesis occurs in response to a broad array of cellular stress conditions, and sphingolipids, including sphingoid bases in yeast and ceramides in mammalian cells, are emerging as key regulators of cell stress responses [12,15,33]. This is best exemplified in S. cerevisiae, where genetics and biochemical techniques have enabled the identification of diverse roles for sphingoid bases in the HSR . These roles include regulation of growth , transcription , protein degradation  and regulation of translation initiation . The current results expand these roles to include formation of p-bodies, as inhibition of de novo sphingolipid synthesis attenuated p-body formation during heat stress. The link between sphingolipid metabolism and p-body formation derives further support from the observations that addition of exogenous sphingoid bases and the use of the lcb4Δ/lcb5Δ double-deletion strain, which shows high constitutive levels of sphingoid bases, results in p-bodies that are readily observed at normal temperature. Thus sphingoid base accumulation appears sufficient to drive p-body formation, and necessary for this process during heat stress.
These results are also consistent with previous findings that sphingoid bases promote translation initiation . Moreover, that study demonstrated that the lcb4Δ/lcb5Δ double-deletion strain, showed elevated rates of translation initiation compared with its parental strain. Together, these results demonstrate a direct correlation between sphingolipids, p-body formation and translation initiation, and suggest that sphingolipids mediate translation initiation by promoting p-body formation. Moreover, the results suggest that a major function of p-bodies during heat stress is to mediate translation initiation.
Thus our results are consistent with a model whereby heat stress precipitates a decrease in translation initiation and an increase in sphingoid bases by separate mechanisms; sphingoid bases are then required, along with translation arrest, for p-body formation. These events then mediate recovery of translation initiation (Figure 6).
Various studies suggest that different stress conditions may regulate p-bodies differently [3,8,9]. Thus whether our model is specific for the HSR remains to be determined. However, these results pose the intriguing possibility that sphingolipids may also mediate translation and/or mRNA movement in higher organisms.
In conclusion, the results from the present study define a novel role for sphingoid bases as signalling molecules required for heat-stress-induced formation of p-bodies, and the results underscore the fundamental role of sphingolipids in the complex regulation of eukaryotic stress responses.
Ashley Cowart, Yusuf Hannun and Jason Gandy, obtained, constructed and/or curated yeast strains and plasmids, cultured yeast, performed some microscopy and measured sphingolipids. Baby Tholanikunnel assisted with the polysome gradient determinations. Ashley Cowart performed microscopy, assisted with the polysome gradients, prepared figures and wrote the manuscript. The study was conceived and directed by and Yusuf Hannun.
This work was supported by the National Institutes of Health [grant number GM63625 (to Y.A.H.)]; the American Heart Association [grant number 0555470U (to B.G.T.)]; the Center of Biomedical Research Excellence in Lipidomics and Pathobiology (to L.A.C.); and by a VA Merit Award (to L.A.C).
We thank Dr Roy Parker, (Howard Hughes Medical Institute, Department of Molecular and Cellular Biology, University of Arizona, Tucson, AZ 85721, U.S.A.) for the gifts of the Dcp2–GFP strain and the plasmid construct for expression of Edc3p–mCherry. Synthetic lipids were provided by Dr Zdzslaw Szulc at the Lipidomics Core Facility at the Medical University of South Carolina. We also acknowledge technical assistance from Sarah Brice. Microscopy was performed in the Medical University of South Carolina Imaging Core facility directed by Dr John Lemasters.
Abbreviations: eIF, eukaryotic initiation factor; GFP, green fluorescent protein; HSR, heat-stress response; mRNP, messenger ribonucleoprotein; p-body, processing body; RFP, red fluorescent protein
- © The Authors Journal compilation © 2010 Biochemical Society