Human cells release Argonaute 14 and major vault protein independently of exosomesAnnexin A1 is a specific marker of microvesicles shed from the plasma membraneSmall extracellular vesicles do not contain DNAActive secretion of cytosolic DNA occurs through an amphisome-dependent mechanism
The heterogeneity of small extracellular vesicles and presence of non-vesicular extracellular matter have led to debate about contents and functional properties of exosomes. Here, we employ high-resolution density gradient fractionation and direct immunoaffinity capture to precisely characterize the RNA, DNA, and protein constituents of exosomes and other non-vesicle material. Extracellular RNA, RNA-binding proteins, and other cellular proteins are differentially expressed in exosomes and non-vesicle compartments. Argonaute 14, glycolytic enzymes, and cytoskeletal proteins were not detected in exosomes. Weidentify annexin A1 as a specific marker for microvesicles that are shed directly from the plasma membrane. We further show that small extracellular vesicles are not vehicles of active DNA release. Instead, we propose a new model for active secretion of extracellular DNA through an autophagy- and multivesicular-endosome-dependent but exosome-independent mechanism. This study demonstrates the need for a reassessment of exosome composition and offers a framework for a clearer understanding of extracellular vesicle heterogeneity.
Synthesis and degradation of cellular constituents must be balanced to maintain cellular homeostasis, especially during adaptation to environmental stress. The role of autophagy in the degradation of proteins and organelles is well-characterized. However, autophagy-mediated RNA degradation in response to stress and the potential preference of specific RNAs to undergo autophagy-mediated degradation have not been examined. In this study, we demonstrate selective mRNA degradation by rapamycin-induced autophagy in yeast. Profiling of mRNAs from the vacuole reveals that subsets of mRNAs, such as those encoding amino acid biosynthesis and ribosomal proteins, are preferentially delivered to the vacuole by autophagy for degradation. We also reveal that autophagy-mediated mRNA degradation is tightly coupled with translation by ribosomes. Genome-wide ribosome profiling suggested a high correspondence between ribosome association and targeting to the vacuole. We propose that autophagy-mediated mRNA degradation is a unique and previously-unappreciated function of autophagy that affords post-transcriptional gene regulation.
Autophagy is a highly conserved eukaryotic pathway that isolates cellular components for degradation and recycling in response to nutrient starvation, thereby maintaining homeostasis in nutrient-limited environments1,2. Upon autophagy induction, substrates are sequestered within a double-membrane vesicle called an autophagosome and subsequently delivered to the vacuole (yeast) or lysosome (mammals) where they are degraded by hydrolytic enzymes.
Autophagy is originally thought to be nonselective (i.e., bulk), which isolates cytoplasmic material in an apparently random manner. In contrast, selective autophagy eliminates harmful proteins and superfluous, damaged organelles, such as endoplasmic reticula (ER) and mitochondria3. Substrates of selective autophagy, often marked for degradation by modifications such as ubiquitination, are recognized through their binding to receptors that facilitate localization to the site of autophagosome formation4,5,6. In addition to the clear distinction of substrate, autophagy may show the broad spectrum of protein substrate preference. Acetaldehyde dehydrogenase and tRNA ligases7,8 are such examples. The mechanism underlying this preference has been still unknown.
Despite the many studies of protein and organelle degradation by autophagy, RNA degradation by autophagy has not been well examined. Historically, RNA degradation by autophagy was suggested in pioneering works examining aminoacid-starved rat livers by Lardeux and Mortimore9,10,11,12. More recently, our group showed that RNA delivered to vacuoles via autophagy is degraded by the T2-type RNase Rny1, and that the resulting nucleotides are further hydrolyzed to nucleosides by the vacuolar phosphatase/nucletidase Pho813. The question of whether RNA degradation by autophagy occurs preferentially, however, remains unaddressed.
In this study, we examine the degradation of mRNAs by rapamycin-induced autophagy in yeast. mRNA sequencing reveals that autophagy preferentially delivers a subset of mRNA species to the vacuole for Rny1-mediated degradation. We further find that the selectivity of autophagy-mediated mRNA degradation is coupled with translation by reporter gene analysis. Genome-wide ribosome profiling further demonstrates that persistent ribosomal interaction is associated with selective mRNA delivery to vacuoles. The autophagic delivery of ribosome-mRNA complexes to the vacuole depends on Atg24 sorting nexin complex, which is required for autophagic degradation of multisubunit complexes, such as the proteasome and ribosome. We provide a foundational description of autophagy-mediated RNA degradation in yeast, as well as molecular insights into the selectivity of this hitherto underappreciated pathway.
We began by identifying mRNAs species delivered to vacuoles by autophagy. As mRNA delivered to the vacuole is immediately degraded by the vacuolar ribonuclease Rny1 in wild-type (WT) cells13, we constructed a strain (rny1) lacking this enzyme. To isolate vacuole-delivered mRNAs, we adopted a strategy employing a highly purified vacuolar fraction from cell lysates by flotation with ultracentrifugation (Fig.1a). The purity of the vacuolar fraction was confirmed by vacuole protein enrichment and the apparent absence of ER, Golgi, and nucleolar proteins (Supplementary Fig.1a).
a Schematic illustration of the experimental design. Following the induction of autophagy by 3h rapamycin treatment, vacuoles from yeasts were purified by ultracentrifugal flotation. A strain deleted for the vacuolar nuclease Rny1, which degrades nucleic acids within the vacuole, was used. b Autophagy-dependent mRNA accumulation in the vacuole fraction. Agarose gel electrophoresis image of RNAs detected in the vacuole fraction of rapamycin-treated (3h) cells. rRNA subtraction was performed by treating extracted RNA samples with oligonucleotides complementary to rRNAs. Representative results from three independent experiments are shown. Asterisk indicates a nonspecific band. c Meta-gene analysis for mRNA-Seq read distributions around the 5 and 3 ends of transcripts from whole-cell lysate (gray) or vacuole fraction (purple). The 5 end of reads is depicted. The line and shaded area represent the median and interquartile range, respectively. d Change in mitochondrial genome-encoded mRNAs enrichment in vacuoles versus total cellular extracts following 3h rapamycin treatment. Box plots show the accumulation of reads originating from nuclear and mitochondrial genomes, as determined by RNA-Seq. The median, interquartile range (IQR), and 1.5 IQR are represented by a solid line, box, and whiskers, respectively. Significance was determined by unpaired two-sided Mann-Whitney U-test.
Using rapamycin to induce autophagy through TORC1 suppression14 and the vacuole purification method described above, we examined the accumulation of RNA species in vacuoles. In addition to rRNAs, we also detected mRNAs that were clearly observed following rRNA subtraction (Fig.1b). Importantly, we determined that the presence of mRNAs in vacuoles depends on autophagy: mRNAs were hardly detected in atg2 cells, which are completely defective for autophagy15.
Next, we characterized the mRNAs in the vacuoles by RNA sequencing. The even distribution of vacuolar mRNA reads suggests that nontruncated, full length mRNAs are delivered to vacuoles in rny1 cells (Fig.1c) and verifies that Rny1 is the sole vacuolar nuclease responsible for mRNA degradation13. In contrast to nuclear genome-encoded mRNAs, mitochondrial genome-encoded mRNAs were clearly excluded from the vacuole fraction (Fig.1d), suggesting only limited mitochondrial degradation under the employed conditions.
We next set out to determine unique features of mRNA species delivered to the vacuole while accounting for previously reported dramatic changes in the cellular transcriptome upon rapamycin treatment (Supplementary Fig.2a and see Materials and Methods)16. A broad spectrum of mRNAs is delivered to the vacuole (Fig.2a and Supplementary Fig.2b). We identified quantitatively over- and under-represented subsets of mRNAs relative to the cellular transcriptome, which we refer to as vacuole enriched and vacuole depleted, respectively. Notably, this analysis further ensured that the delivery of vacuole-enriched mRNAs is dependent on Atg2 (Fig.2b). Northern blotting of representative vacuole-enriched mRNAs further confirmed autophagy-dependent mRNA enrichment in the vacuole (Supplementary Fig.2c).
a Relative rapamycin-induced mRNA delivery into vacuoles. MA (log ratio vs. mean average) plot of 5592 mRNAs showing relative mRNA enrichment in vacuoles following rapamycin treatment (3h) versus transcript per million (TPM) in rny1 cells in growing condition. mRNAs identified as deviating significantly from total cell extracts were classified as enriched (log2-fold change1 and q value<0.01) or depleted (log2-fold change1 and q value<0.01), as indicated in red and blue, respectively. b Autophagy dependency of vacuole-enriched mRNA delivery to the vacuole. Box plot showing the relative mRNA enrichment of vacuole-enriched mRNAs following rapamycin treatment (3h) in rny1 and rny1 atg2 cells. The median, IQR, and 1.5 IQR are represented by solid line, box, and whiskers, respectively. Significance was determined by unpaired two-sided Mann-Whitney U-test. c Gene ontology analysis of mRNAs by relative vacuolar enrichment, as determined by iPAGE62. The representative vacuole-enriched (red) or -depleted (blue) genes analyzed in this study are shown. d Time course of vacuole-enriched (red) or -depleted (blue) mRNA accumulation in rny1 cell vacuoles following rapamycin treatment. e Overview of a labeling strategy employing 4-thiouracil (4-thioU) with rapamycin treatment to monitor mRNA degradation. f Persistence of 4-thioU-labeled vacuole-enriched (red) or -depleted (blue) mRNAs following 1h rapamycin treatment as determined by the strategy shown in e. Data were normalized to 4-thioU-labeled mRNA values detected in WT cells, which were set to 1. In d and f, data present mean (line) and individual results (points) of three independent experiments.
Next, we conducted functional characterization of the vacuole-enriched and -depleted mRNAs by gene ontology analysis. Strikingly, housekeeping mRNAs associated with amino acid biosynthesis and ribosomal proteins were most likely to be delivered to vacuoles (Fig.2c). In contrast, the vacuole-depleted fraction was characterized by mRNAs with regulatory roles, such as those encoding proteins with protein kinase, energy reserve metabolism, and transcriptional activator (Fig.2c). We note that the mRNA selectivity is independent of basal mRNA abundance in cells (Fig.2a). Taken together, these results reveal that autophagy-mediated mRNA delivery to vacuoles is selective, not random, in nature.
We quantified the proportion of mRNA delivered to the vacuole. To this end, we used the activity of vacuolar alkaline phosphatase17,18 as a standard to ascertain the recovery of vacuoles from cell lysates and thereby the relative quantity of mRNAs delivered to the vacuole. This strategy revealed that for vacuole-enriched mRNA species, the proportion of mRNAs accumulating in the vacuole increased dramatically upon autophagy induction (Fig.2d and Supplementary Fig.2c); more than 20% of vacuole-enriched HOM2 (aspartic beta semi-aldehyde dehydrogenase) and LYS1 (saccharopine dehydrogenase) mRNAs in cell lysates were delivered to vacuoles following 3h of rapamycin treatment. In contrast, the fraction of vacuole-depleted species, such as APL5 (a subunit of the clathrin associated protein complex) and ASG1 (zinc cluster protein proposed to be a transcriptional regulator), was limited at most to a 5% increase in vacuolar localization (Fig.2d and Supplementary Fig.2d).
As vacuole-enriched mRNAs are preferentially delivered to the vacuole, we reasoned that the relatively high rate of degradation of these mRNAs should be observed in WT cells. This was verified by RNA-Seq analyses of WT total cell lysates: the total abundance of vacuole-enriched mRNAs was significantly decreased following rapamycin treatment (Supplementary Fig.2e, f). To more directly monitor mRNA degradation, we labeled cellular mRNAs with 4-thiouracil (4-thioU) in vivo, after which we subjected cells to chase in uracil-containing media (Fig.2e). For this experiment, we used ura3 cells to enhance incorporation of modified uracil into mRNAs (Supplementary Fig.3a). In addition, we ascertained that the deletion of ATG2 or RNY1 does not affect the RNA labeling rate and total mRNA levels with 4-thioU (Supplementary Fig.3a, b). Following 1h rapamycin treatment, labeled mRNAs were biotinylated and purified using streptavidin beads, followed by quantification of mRNA abundance. As the reliability of short-lived mRNA analysis is poor, we focused on relatively long-lived mRNAs19. Two representative vacuole-enriched mRNAs engaged in amino acid biosynthesis, HOM2 and HIS5 (histidinol-phosphate aminotransferase), were detected at between 3- to 4-fold abundance in atg2 cells in comparison to WT cells (Fig.2f, top), providing further evidence of autophagy-dependent degradation. Remarkably, even with an intact autophagy machinery, vacuole-enriched mRNAs in rny1 cells were also retained to a nearly identical degree as that observed in atg2 cells, indicating that the nuclease activity of Rny1 is responsible for the degradation of mRNAs following delivery to the vacuole. Meanwhile, vacuole-depleted APL5 and NTH2 (putative neutral trehalase) mRNAs showed only marginal autophagy-dependent differences in stability (Fig.2f, bottom). These results highlight the impact of autophagy on mRNA degradation and clearly demonstrate that autophagic mRNA degradation is carried out via preferential vacuole delivery and subsequent Rny1-mediated hydrolysis.
As for other forms of selective autophagy in protein degradation, the mRNA degradation pathway must employ a distinct mechanism. By comparing vacuolar mRNA delivery and mRNA stability, we did not find the correspondence between selectivity of mRNA delivery to the vacuole and the steady-state half-life of mRNAs19 (Supplementary Fig.4a). Thus, the recognition and isolation of target mRNAs necessary for vacuolar degradation appeared to be different to canonical cytoplasmic mRNA turnover, leading us to investigate the mechanistic basis of vacuolar mRNA degradation.
Selective degradation of the ER has been reported20,21, and many mRNAs are bound to the ER surface22. We determined that concomitant degradation of ER-bound mRNAs is unlikely as ER-bound mRNAs were not enriched in vacuoles (Supplementary Fig.4b). Another possibility is that mRNAs within ribonucleoprotein (RNP) granules, such as stress granules, are degraded by autophagy through their association with RNP granules23,24. However, stress granule mRNAs25 were not enriched in the vacuole; rather, these mRNAs were predominantly identified in the vacuole-depleted fraction, suggesting that stress granules are not degraded by autophagy under the conditions employed in this study (Supplementary Fig.4c).
Autophagy is also known to sequester cytoplasmic ribosomes into autophagosomes for degradation in the vacuole26. We observed the accumulation of rRNA in the vacuolar fraction following autophagy induction (Fig.3a), giving rise to the possibility that translating mRNAs are delivered to the vacuole. To test this hypothesis, we designed a reporter assay to monitor mRNA enrichment by using the vacuolar delivery of HOM2 mRNA as a model (Fig.2c, d, and Supplementary Fig.2c, d). We generated and expressed a series of reporter plasmids under the control of the native HOM2 promoter (ensuring endogenous expression levels) in a hom2 background strain (Fig.3b). Expression of WT HOM2 mRNA in this strain recapitulated vacuolar delivery. In contrast, delivery of reporter mRNA to the vacuole was inhibited by the insertion of a strong stem-loop just before the start codon, which blocks start codon scanning by the pre-initiation 43S complex and thus impedes translation27,28 (Fig.3b, c). Further, blocking of translation by the elimination of the start codon (substitution to TTG or TAC) also reduced mRNA delivery into vacuoles while inhibiting translation (Fig.3b, c). Reporter mRNA level in total cell lysate were comparable with WT HOM2 mRNA (Supplementary Fig.5d). Alternative vacuole-enriched ARO2 (chorismite synthase/flavin reductase) mRNA, which was similarly engineered as HOM2 reporter, also showed translation dependency for vacuolar delivery (Supplementary Fig.5a, b, e). These results suggest that mRNA delivery to the vacuole is closely coupled with translation.
a Autophagy-dependent rRNA accumulation in the purified vacuole fraction. Agarose gel electrophoresis of RNAs from purified vacuoles with or without 3h rapamycin treatment. Representative results from three independent experiments are shown. b, d Proportion of a representative vacuole-enriched reporter mRNA, HOM2, recovered from vacuolar fractions following 3h rapamycin treatment. Data were analyzed as described in Fig.2d. In b, translation was blocked either through the introduction of an inhibitory stem-loop just before the HOM2 start codon, or by introducing point mutations into the start codon (ATG to TTG or TAC). In d, the promoter-5 UTR region of HOM2 was replaced with that of vacuole-depleted APL5 or ASG1. Data present mean (line) and individual results (points) of three independent experiments. c, e Protein levels of reporter mRNAs used in b and d as determined by Western blotting in growing condition. Representative results from two independent experiments are shown.
To determine whether a protein-encoding region of vacuole-enriched mRNAs is responsible for autophagic delivery to the vacuole, we swapped the HOM2 ORF with an ORF encoding GFP. Substitution of the HOM2 ORF had little effect on vacuolar delivery (Supplementary Fig.5c). In addition, the substitution of the 3 UTR of HOM2 with that of PIG2 (putative type-1 protein phosphatase targeting subunit), a representative vacuole-depleted mRNA, had only a very weak effect on vacuolar delivery (Supplementary Fig.5c, f). We next swapped the HOM2 5 UTR with vacuole-depleted mRNA 5 UTRs. We employed 5 UTRs from vacuole-depleted mRNAs expressed at a similar mRNA level to WT HOM2 mRNA, since a low abundance of reporter mRNAs could result in the overestimation of vacuolar delivery (Supplementary Fig.5g). The substitution of the HOM2 5 UTR with vacuole-depleted APL5 or ASG1 5 UTRs clearly reduced the efficiency of delivery (Fig.3d). Importantly, the 5 UTR of vacuole-depleted mRNAs was associated with a marked decrease in HOM2 protein level (Fig.3e), which is consistent with the coupling of translation and degradation. Thus, we propose that the 5 UTR is critical for vacuolar delivery through its regulation of vacuole-enriched mRNA translation.
We next conducted an analysis of ribosome-mRNA interactions at the genome-wide level. To this end, we performed ribosome profiling29 and monitored associations upon rapamycin treatment, normalizing data to mRNA abundances obtained by RNA-Seq. Rapamycin treatment reduced global protein synthesis, as indicated by reduced polysomes formation (Supplementary Fig.6a) and previously reported30. Remarkably, ribosomal association with vacuole-enriched HOM2 and LYS1 mRNAs persisted during rapamycin treatment (Fig.4a). In contrast, vacuole-depleted NTH2 and APL5 mRNAs exhibited a time-dependent decrease in ribosomal association (Fig.4a). These trends were observed with a high degree of reproducibility throughout all vacuole-enriched and -depleted transcripts (Fig.4b) and at all assessed time points following rapamycin treatment (1, 2, and 3h) (Supplementary Fig.6b). Taken together, these results strongly suggest that persistent ribosome-mRNA association is a key determinant of mRNA degradation by autophagy, even as global translation activity is inhibited during rapamycin treatment.
a Ribosome association (ribosome footprint data normalized to RNA-Seq reads) of the vacuole-enriched (red) or -depleted (blue) mRNAs (defined in Fig.2a) during rapamycin treatment. b Cumulative distribution of vacuole-enriched (red) and -depleted mRNAs (blue) (defined in Fig.2a) in relation to the change in ribosome association following 3h of rapamycin treatment. Significance was determined by unpaired two-sided Mann-Whitney U-test. c Agarose gel electrophoresis of RNAs recovered from whole-cell lysate or purified vacuoles of rny1 and rny1 atg24 cells following 3h rapamycin treatment. Representative result from three independent experiments is shown. d The vacuole-enriched (red) or -depleted (blue) mRNA accumulation in purified vacuoles of rny1, rny1 atg24, rny1 atg20, and rny1 snx41 cells. Data were analyzed as described in Fig.2d. Data present mean (line) and individual results (points) of three independent experiments. e A model for autophagy-mediated mRNA degradation in yeast. A subset of mRNAs is preferentially delivered to vacuoles by autophagy. This mRNA delivery is coupled to mRNA translation. Persistence of ribosome association with mRNA enhances selective mRNA delivery to vacuoles even during blockage of global protein synthesis by TORC1 inhibition. mRNAs delivered to the vacuole by autophagy are subsequently degraded by the nuclease Rny1.
A previous study reported that Atg24 (also known as Snx4) is specifically required for the autophagic degradation of multisubunit complexes such as ribosomes, but not of bulk proteins31. Given that autophagic degradation of mRNA is coupled to ribosomal association, we hypothesized that Atg24 is required for autophagy-mediated RNA degradation. To test this, we assessed autophagic degradation of GFP-fused proteins by western blotting, whereby autophagic delivery to the vacuole and subsequent degradation of a protein is detected by the liberation of the vacuolar protease-resistant GFP moiety. As reported previously31, the cleavage of free GFP from the bulk-autophagy reporter Pgk1-GFP was observed in both atg24 and WT cells (Supplementary Fig.7a), whereas cleavage of GFP from the ribosomal protein Rpl37a-GFP was detected in WT but not atg24 cells (Supplementary Fig.7b). Critically, accumulation of rRNAs in the vacuoles of rny1atg24 double mutant cells was not observed (Fig.4c), indicating that ribosomal degradation by autophagy is largely dependent on Atg24. Quantification of the amount of mRNA delivered to the vacuole suggests that vacuolar mRNA delivery was severely inhibited in the absence of Atg24 following rapamycin treatment (Fig.4d).
Atg24 consists of a sorting nexin complex with Atg20 or Snx4132,33,34. We also tested the impact of Atg20 and Snx41 for vacuolar delivery of mRNAs. Clearly, the deletion of these genes reduced the efficiency of mRNA delivery to the vacuole (Fig.4d), suggesting that Atg24 functions as complexes with Atg20 or Snx41 for autophagy-mediated mRNA delivery into vacuoles. Overall, we conclude that ribosome-bound mRNAs are preferentially delivered to vacuoles by the autophagy machinery and facilitate their degradation (Fig.4e).
Whereas cytoplasmic RNA degradation pathways, including canonical mRNA decay and the mRNA surveillance system, have been well studied, vacuolar degradation of mRNAs mediated by autophagy remains poorly understood. In this study, we demonstrate the delivery of mRNAs to vacuoles by autophagy, and that this process is mRNA-selective in nature (Fig.4e). Following TOR inhibition, a subset of mRNAs, including those implicated in amino acid biosynthesis and ribosomal protein-encoding mRNAs, are delivered by rapamycin-induced autophagy to the vacuole, where they are degraded by the vacuolar nuclease Rny1. The delivery of mRNAs requires Atg24 and is enhanced by their ribosomal association, the latter of which is mediated by the 5 UTR. The negative correlation between vacuole delivery and stress granule enrichment (Supplementary Fig.4c) may reflect the relatively low translation rate of mRNAs within stress granules, thereby likely inhibiting autophagy-mediated vacuolar delivery.
Considering that mRNAs are apparently delivered to the vacuole in the form of a polysome, the selective degradation of ribosomes by autophagy (ribophagy)35 offers one potential explanation for the degradation of mRNAs observed in this study. Previous works reported that ribophagy occurs following a long period (~24h) of nitrogen starvation35,36,37,38. However, in this study we observe the delivery of mRNAs to vacuoles at much earlier time points (at 3h) by chemical TORC1 inhibition (Fig.2)13. Moreover, while ribophagy has been reported to depend on the Ubp3/Bre5 deubiquitination complex35, RNAs are still delivered into vacuoles even in the absence of these factors13. Although the relation between translation status and ribophagy has not been well studied, the preferential mRNA degradation that we found in this study cannot be attributed to previously described ribophagy35. While the clear sequestration of ribosomes within vacuoles has been reported26, whether mammalian ribophagy occurs is currently a point of contention. While one study reported containing ribosomes within autophagosomes and attributed their sequestration into autophagosomes to the receptor protein NUFIP139, another showed only limited degradation of ribosomes by autophagy during nutrient stress40. Meanwhile, the proportion of ribosomal protein in yeast is an order of magnitude higher than that of mammalian cells41. Ribophagy may therefore differently contribute to ribosomal turnover and mRNA degradation in various species.
There are two potential explanations for persisted ribosome-mRNA association during rapamycin treatment: increased translation initiation or reduced translation elongation, with the former being the most straightforward interpretation of the data presented in this study. In the latter case, reduced translation elongation would result in a high representation of ribosome footprints proximal to the 5 end of ORFs42. We assessed the position of ribosome association on mRNAs by calculating polarity score42, but did not observed distribution of ribosome footprints proximal to the 5 end of ORFs in vacuole-enriched mRNAs (Supplementary Fig.8a). Enhanced translation initiation is also consistent with the important role of 5 UTRs in translation: these regions possess important regulatory elements that determine the rate of translation initiation43. The upstream ORF (uORF), a region found within the 5 UTR of genes such as GCN4 (transcriptional activator of amino acid biosynthetic genes)44,45, allows preferential translation under stress46 and may support such enhanced translation initiation (Supplementary Fig.8b, c). However, ribosome profiling did not identify any uORFs among vacuole-enriched mRNAs (Supplementary Fig.8d). We only observed that small differences in the length of 5 UTR47 between vacuole-enriched and vacuole-depleted mRNAs (Supplementary Fig.9a, b). RNA binding protein(s) that interact with the 5 UTR of implicated mRNAs may also facilitate recognition by the autophagy machinery, although we did not find any consensus sequences by motif analysis of vacuole-enriched mRNAs.
Another question is how delivered translating ribosomes or ribosome-mRNA complexes (polysomes) are preferentially recognized by the autophagy machinery. Interaction with adaptor proteins, as observed in selective autophagy4,5,6, or specific, transient protein/RNA modifications48,49 may act as intermediaries that recruit the autophagy machinery for mRNA degradation. In any case, our data imply that the state of polysomes in the cytoplasm is not homogeneous and dynamically changes under cellular conditions.
In this study, we induced autophagy using rapamycin, a pharmacological inhibitor of TORC1 widely used in autophagy research, but the investigation of mRNA degradation selectivity under other conditions promises to shed further light on interplay between ribosomal association and mRNA delivery. Using this strategy to understand the global landscape of vacuole-enriched mRNAs under a range of autophagy-inducing conditions may provide a generalized rationale for mRNA preference.
We propose that autophagy acts as an mRNA degradation system at the translation step, adding a mechanism for transcriptional regulation and global translation inhibition during cellular adaptation to stress conditions. The autophagic degradation of ribosome-associated mRNAs in the early phase of the stress response may be required to facilitate the shift to translation of stress response genes. Degradation of mRNAs engaged in ribosome association may be counter-intuitive, as this may suggest inefficient adaptation to stress when the cell is least able to tolerate the waste of resources. However, such degradation likely reflects an intricate balance between the expression of target genes and the inhibition of needless expression of target mRNAs. This study demonstrates that autophagy is involved in the regulation of gene expression through mRNA degradation, thereby playing a critical role in cellular homeostasis.
Strains used in this study are listed in Supplementary Table1. Gene disruption and tagging were performed using a standard PCR-based method50,51. Cells were grown in rich medium (1% yeast extract, 2% peptone, and 2% glucose) or synthetic defined casamino acid medium without uracil (SDCA-uracil: 0.17% Difco yeast nitrogen base without amino acids and ammonium sulfate, 0.5% casamino acids, 0.5% ammonium sulfate, 0.002% tryptophan, 0.002% adenine, and 2% glucose). Yeast cells were grown in liquid media at 30C to a density of OD600=1.0 before autophagy was induced by the addition of 0.2M rapamycin (LC Laboratories, R-5000). The strains prepared in this study will be distributed upon request.
Yeast vacuoles were isolated from whole cells as previously described26,52 with some modifications. Cells were grown in rich medium or SDCA-uracil to a density of OD600=1.0 before supplementation of 0.2M rapamycin and incubation for further 1, 2, or 3h. Spheroplasts were prepared by incubation of cells in spheroplast buffer (1.2M sorbitol, 50mM Tris-HCl pH 7.5, 50mM 2-mercaptoethanol, and 5 U/ml Zymolyase 100T [nacalai tesque, 07665-55]) for 30min. Spheroplasts were then collected by centrifugation, washed with wash buffer (1.2M sorbitol and 50mM Tris-HCl pH 7.5), resuspended in ice-cold buffer A (12% w/v Ficoll 400, 0.1mM MgCl2, 10mM MES-Tris pH 6.9, and 1 complete, EDTA-free protease inhibitor cocktail [Roche, 5056489001]), and homogenized with Dounce homogenizer on ice. All the subsequent handling was performed on ice. The lysate was next transferred to an ultracentrifuge tube and buffer B (8% w/v Ficoll 400, 0.1mM MgCl2, and 10mM MES-Tris pH 6.9) was layered on top, and centrifuged at 72,000g in a P28S swinging bucket rotor (Hitachi Koki) for 30min at 4C. The top, white layer (crude vacuole) was collected in new ultracentrifuge tube and resuspended in buffer B. The crude vacuole was underlaid with buffer B (4% w/v Ficoll 400, 0.1mM MgCl2, and 10mM MES-Tris pH 6.9) and subjected to a further round of centrifugation at 72,000g in a P40ST swinging bucket rotor (Hitachi Koki) for 30min at 4C. The top layer was collected (the vacuole fraction) and RNA or protein were isolated immediately. The vacuole fraction used in analysis in Figs.2d, 3, 4c, 4d, and Supplementary Fig.5 was treated with RNase I (Lucigen, N6901K). The RNase I was inactivated by dithiothreitol (DTT) for 20min at 70C before RNA isolation.
The activity of a vacuolar enzyme, alkaline phosphatase (ALP), was employed as previously described53 to biochemically determine the yield of vacuoles recovered from cells by ultracentrifugation. Five microliter of vacuole fraction or whole-cell lysate were suspended in 500l of assay buffer (250mM Tris-HCl pH 9.0, 10mM MgSO4, and 10M ZnSO4). The reaction was initiated by addition of 50l of 55mM -naphthyl phosphate. After incubation for 10min at 30C, 500l of 2M glycine-NaOH (pH 11.0) was added to stop the reaction. The fluorescence intensity was measured at 345nm excitation and 472nm emission. The vacuole recovery was calculated by the fluorescence intensity of the vacuole fraction normalized to that of whole-cell lysate.
RNAs from yeast lysate and isolated vacuoles were extracted by TRIzol reagent (Thermo Fisher Scientific, 15596018) following the manufacturers instructions. rRNA subtraction was performed using the Ribominus Transcriptome Isolation Kit (Yeast) (Invitrogen, K155003) according to the manufacturers instructions. For electrophoresis, the extracted RNAs were separated on denaturing formaldehyde agarose gel (1% agarose, 1 MOPS buffer pH 7.0 [20mM MOPS, 5mM NaOAc, and 1mM EDTA], and 2% formaldehyde) and stained with GelRed (1:3300 dilution in water) (Biotium, 41003). For Northern blotting, digoxigenin (DIG) conjugated probes were prepared by PCR DIG Probe Synthesis Kit (Roche, 11636090910) according to the manufacturers instructions. The probe was hybridized using PerfectHyb Plus (SigmaAldrich, H7033) for 2h at 68C. A DIG-labeled HOM2 probe was prepared from isolated yeast genomic DNA using 5-CGTTGGTCAACGTTTCATTCTGTTGTTG-3 and 5-AGTGGTCAAAGCATCAATAGGACCG-3 oligonucleotides. DIG-labeled ARO2 probe was prepared using 5-CACCACATATGGTGAATCGCATTGTAAGT-3 and 5-GTTCAACAGATGCTGAAATTCAGGATCG-3 oligonucleotides. Anti-digoxigenin-AP (Roche, 11093274910, 1:10000) and CDP-Star (Roche, 11685627001) were used for signal detection. Chemiluminescence images were acquired using a FUSION-FX7 (Vilber-Lourmat) imaging systems. For qPCR, cDNAs were prepared using the PrimeScript RT reagent kit with gDNA eraser (TAKARA, RR047). A random hexamer was used for cDNA synthesis. Subsequent qPCR was conducted using TB Green Premix Ex Taq II (Tli RNase H Plus) (TAKARA, RR820) with the mRNA-specific primers listed in Supplementary Table3. Serial dilutions of cDNA were used for qPCR calibrations. Melting-curve analyses confirmed the amplification of a single product for each mRNA. The proportion of each mRNA species in the vacuole (the proportion in vacuoles) was calculated by the value of the vacuole fraction normalized to that of the total lysate, considering the vacuole recovery rate as described above.
Two hundred nanogram of RNA was isolated from total cell lysates or vacuolar fractions using TRIzol reagent (Thermo Fisher Scientific, 15596018) for extraction and and Direct-zol RNA MiniPrep Kit (Zymo Research, R2052) for purification, both according to the manufacturers instructions. After rRNA depletion was performed using the Ribo-Zero rRNA Removal Kit (yeast) (Illumina, MRZY1324), libraries were prepared using the TruSeq RNA Library Preparation Kit v2 (Illumina, RS-122-2001), according to the manufacturers instructions. The multiplexed libraries were sequencing on a HiSeq 4000 sequencer (Illumina) for single-end 50bp.
Thirty OD600 units of yeast cells treated with 0.2M rapamycin for 1, 2, and 3h or without rapamycin treatment (0h) were collected by filtration. Ribosome profiling libraries were prepared by modifying a previously described method54,55. The collected cells with drip of ribosome profiling lysis buffer (20mM Tris-HCl pH 7.5, 150mM NaCl, 5mM MgCl2, 1mM DTT, 1% Triton X-100, 100g/ml cycloheximide [SigmaAldrich, C-7698], and 100g/ml chrolamphenicol [Wako, 032-19451]) were ground using Multi-Beads Shocker (Yasui Kikai) with chamber and ball precooled with liquid nitrogen. The cell lysate was centrifuged at 3000g for 5min to remove cellular debris. The supernatant was collected to a fresh tube and centrifuged at 20,000g for 10min to remove nuclei and other cellular debris. The supernatant containing 20g of total RNA was treated with 0.5 U/g of RNase I (Lucigen, N6901K) at 25C for 45min. A linker oligonucleotide 5-(Phos)NNNNNIIIIIAGATCGGAAGAGCACACGTCTGAA(ddC)-3, where (Phos) and (ddC) indicate 5 phosphorylation and a terminal 2, 3-dideoxycytidine, respectively, was used. The Ns and Is indicate random barcode (i.e., unique molecular index, UMI) for eliminating PCR duplication and multiplexing barcode, respectively. The linkers were preadenylated with 5 DNA Adenylation Kit (NEB, E2610S) and then purified using Oligo Clean & Concentrator (Zymo Research, D4060). The linker ligated footprints were reverse transcribed with a primer 5-(Phos)NNAGATCGGAAGAGCGTCGTGTAGGGAAAGAG(iSp18)GTGACTGGAGTTCAGACGTGTGCTC-3, where iSp18 stands for an internal 18-atom hexa-ethyleneglycol spacer, and circularized by CircLigase II (Lucigen, CL9025K). The cDNA was PCR-amplified using primers 5- AATGATACGGCGACCACCGAGATCTACACTCTTTCCCTACACGACGCTC-3 and 5-CAAGCAGAAGACGGCATACGAGATATCACGGTGACTGGAGTTCAGACGTGTG-3.
For library preparation of RNA-Seq, total RNA was extracted from the same lysate for ribosome profiling using TRIzol LS reagent (Thermo Fisher Scientific, 10296010) and Direct-zol RNA MiniPrep Kit (Zymo Research, R2052). Libraries were prepared using the TruSeq RNA Library Preparation Kit v2 (Illumina, RS-122-2001) as described in the product documentation.
The adapter sequence was trimmed using the FASTX_clipper, a part of the FASTX-Toolkit (https://www.scirp.org/(S(351jmbntvnsjt1aadkposzje))/reference/ReferencesPapers.aspx?ReferenceID=1037549). Bowtie256,57 was used to map the clipped reads to yeast rRNAs and capture unaligned reads. The unaligned reads were mapped to the S288C S. cerevisiae reference genome using Tophat58 and counted by HTSeq59. All genes in Saccharomyces Genome Database (SGD) (https://www.yeastgenome.org)60 with reliable sequence annotations and 10 read counts were analyzed by DESeq61 to calculate vacuolar fraction and whole-cell reads. mRNA enrichment in vacuoles was calculated with DESeq using a generalized linear model to normalize overall mRNA expression change in total lysate after rapamycin treatment. Vacuole-enriched and -depleted mRNAs were defined as transcripts with accumulation in vacuoles at more or less than two-fold vacuolar enrichment, respectively, and with a q value below 0.01. Gene ontology analysis was performed using the iPAGE62 web interface (https://tavazoielab.c2b2.columbia.edu/iPAGE/).
Reads mapped to yeast noncoding RNAs (rRNAs, tRNAs, snRNAs, snoRNAs, and mitochondrial rRNAs) were excluded. The remaining reads were then aligned to the S288C yeast genome with STAR63. PCR-duplicated reads with the same UMI in the linker sequence were removed from downstream analysis. Footprints with 2630 nt were used for analysis. The distance of the A-site from the 5-end of the reads was estimated for each footprint length as 15 nt for 2628 nt reads and 16 nt for 29 and 30 nt reads. The change in ribosome association on mRNAs throughout time course of rapamycin treatment was calculated with DESeq using a generalized linear model to normalize overall mRNA expression change after rapamycin treatment. Polarity scores were calculated as a read distribution bias along CDS, as previously described42. The reads located at the A-site in first and last 15 nt of CDS were excluded from analysis. Genes with 64 read counts in CDS were analyzed.
A PCR fragment encoding Renilla reniformis luciferase was amplified from the psiCHECK plasmid using 5-TAATACGACTCACTATAGG-3 and 5-CACACAAAAAACCAACACACAG-3 oligonucleotides (Supplementary Table3). RNA was transcribed from this fragment using the T7-Scribe Standard RNA IVT Kit (CELLSCRIPT, C-AS2607) and 4-thioUTP. After transcription, a cap structure and poly(A) tail were added using the ScriptCap m7G Capping System (CELLSCRIPT, C-SCCE0625) and A-Plus Poly(A) polymerase Tailing Kit (CELLSCRIPT, C-PAP5104H), respectively.
The 4-thiouracil (4-thioU) metabolic labeling and purification of 4-thioU-labeled RNA was performed by modifying a previously described method64. Briefly, cells were grown in synthetic defined medium with low concentration of uracil (SD-low uracil: 0.2% Yeast Synthetic Drop-out Medium Supplements without uracil [SigmaAldrich, Y1501], 0.17% Difco yeast nitrogen base without amino acids and ammonium sulfate, 0.0013% uracil, and 2% glucose) overnight (yielding a density of OD600 24). Next, cells were diluted into the SD-low uracil medium at OD600 0.1. Upon growth of cells to OD600 0.6, cells were incubated in the presence of 1mM of 4-thioU (SigmaAldrich, 440736) for 2h before washing out into fresh synthetic defined medium (SD: 0.2% Yeast Synthetic Drop-out Medium Supplements without uracil [SigmaAldrich, Y1501], 0.17% Difco yeast nitrogen base without amino acids and ammonium sulfate, 0.002% uracil, and 2% glucose) without 4-thioU. Cells were then treated with rapamycin for 1h and collected immediately.
RNAs were extracted by a hot phenol method, as described previously13. Frozen cells were resuspended in 400l of AE buffer (50mM sodium acetate and 10mM EDTA pH 5.0) with 1% SDS, and placed at 65C. Immediately, cells were resuspended in 500l of AE buffer-saturated hot phenol and homogenized using 0.5mm zirconia beads (Yasui Kikai, YZB05) and FastPrep-24 (MP Biomedicals, 6004-500) with a setting of 30s at 5.5m/s for four times. The samples were centrifuged at 15,000g for 10min and the RNA were extracted with ANE buffer (10mM sodium acetate, 100mM NaCl, and 2mM EDTA)-saturated phenol:chloroform and then chloroform:isoamyl alcohol. Subsequently the RNA was precipitated with ethanol. Ten microgram of total RNA was mixed with 5ng of in vitro transcribed 4-thioU-labeled Renilla reniformis luciferase mRNA, biotinylated using 0.2mg/ml of EZ-Link HPDP-Biotin (Thermo Fisher Scientific, 21341) in 10mM Tris-HCl pH 7.5 and 1mM EDTA for 2h at 23C in the dark, and then precipitated with isopropanol. The biotinylated RNAs were incubated with Dynabeads MyOne Streptavidin C1 (Thermo Fisher Scientific, DB6500) for 15min. The beads were washed with buffer 1 (100mM Tris-HCl pH 7.4, 0.5M EDTA, and 5M NaCl) prewarmed to 65C once, buffer 2 (100mM Tris-HCl pH 7.4, 0.5M EDTA, and 10% SDS) once, and then 10% buffer 1 twice. The RNAs were eluted from the streptavidin beads with 5% -mercaptoethanol for 5min at room temperature and then for 10min at 65C, and subsequently precipitated with isopropanol.
4-thioU incorporation into RNA was evaluated by a dot-blot assay, as previously described65 with some modifications. Biotinylated RNA was spotted onto the positively charged nylon membrane and then crosslinked to the membrane by exposure to UV light. The membrane was incubated with 10 blocking solution (125mM NaCl, 9mM Na2HPO4, 7mM NaH2PO4, and 10% SDS) for 20min. Anti-streptavidin-HRP (Abcam, ab7403, 1:5000) was added to the solution and incubated for 10min. The membrane was washed twice by 1 blocking solution for 10min and twice by wash solution (10mM Tris-Base, 10mM NaCl, and 1.05mM MgCl2, pH 9.5) for 5min. Chemiluminescence images detected with Femtoglow HRP Substrate (Michigan Diagnostics, 21008) were acquired by a FUSION-FX7 (Vilber-Lourmat) imaging system.
Western blotting was performed with 0.2 OD600 units of cells harvested for sample preparation. Frozen cells were treated with 10% trichloroacetic acid (TCA) for 5min on ice. After TCA was discarded by centrifugation, the pellet was washed with ice-cold acetone and resuspended in 1 sample buffer (75mM Tris-HCl pH 6.5, 10% glycerol, 25mM DTT, and 0.6% SDS). The samples were homogenized using 0.5mm zirconia beads (Yasui Kikai, YZB05) and FastPrep-24 (MP Biomedicals, 6004-500) with a setting of 60s at 6.0m/s, and then incubated at 65C for 10min. Samples were separated by SDS-PAGE followed by Western blotting. Anti-FLAG (SigmaAldrich, F3165, 1:1000), anti-Pho8 (Abcam, ab113688, 1:1000), anti-Ape1 (1:5000)66, anti--actin (Wako, 010-27841, 1:500), anti-Dpm1 (Invitrogen, A6429, 1:1000), anti-Gsp1 (ImmuQuest, IQ241, 1:10000), anti-Van1 (a gift from Koji Yoda, 1:3000), and anti-GFP (Roche, 11814460001, 1:1000) were used as primary antibodies. Chemiluminescence was raised by Femtoglow HRP Substrate (Michigan Diagnostics, 21008) and blots were visualized using LAS-4000 (GE Healthcare) or FUSION-FX7 (Vilber-Lourmat) imaging systems.
Two hundred OD600 units of yeast cells grown in rich medium with or without rapamycin were treated with 0.1mg/ml of cycloheximide (SigmaAldrich, C-7698) for 5min on ice and collected by centrifugation. The cell pellet was flash-frozen in liquid nitrogen, ground with a mortar and pestle precooled on liquid nitrogen, and then resuspended in lysis buffer (20mM HEPES-KOH pH 7.4, 100mM potassium acetate, and 2mM magnesium acetate). The cell lysate was centrifuged at 1100g for 10min to remove cellular debris. The supernatant was collected to a fresh tube and centrifuged at 9100g for 10min to remove nuclei and other cellular debris. Supernatants containing 100g of total cellular RNA were layered on top of sucrose gradients (1050% sucrose in 10mM Tris-acetate pH 7.4, 70mM ammonium acetate, and 4mM magnesium acetate) prepared in ultracentrifuge tubes (Hitachi Koki) using a Gradient Master (BIOCOMP). Samples were then centrifuged at 150,000g in a P40ST rotor (Hitachi Koki) for 2.5h at 4C. Gradients were fractionated using Piston Gradient Fractionator (BIOCOMP). Continuous absorbance was measured at 254nm using a single path UV-1 optical unit (Biomini UV-monitor, ATTO).
DNA fragments containing the HOM2 or ARO2 genes spanning a region from 500 nt upstream of the initiation codon (containing promoter and 5 UTR sequences) to the 3 UTR were amplified by PCR from genomic DNA. These DNA fragments were inserted to pRS416-CYC1t URA3 CEN using SacI and XhoI sites. A FLAG-tag encoding sequences was then inserted into the region upstream of stop codon by inverse PCR.
A DNA fragment containing BamHI and SalI sites were inserted into 10 nucleotides upstream of the start codon of HOM2 in pRS416-HOM2-FLAG-CYC1t by inverse PCR. Two DNA oligonucleotides (5-GATCCCCCGGAGATCCCGCGGTTCGCCGCGGGCGTACG-3 and 5-TCGACGTACGCCCGCGGCGAACCGCGGGATCTCCGGGG-3) were annealed and then inserted into the pRS416-HOM2-FLAG-CYC1t using BamHI and SalI sites. See also Supplementary Table3 for the list of primers used for this construction.
A DNA fragment containing the ORF and 3 UTR of ARO2 was amplified by PCR from genomic DNA and then inserted into pRS416-stem-loop-HOM2-FLAG-CYC1t using SalI and XhoI sites. A PCR-amplified DNA fragment containing the 500 nucleotides upstream of the ARO2 initiation codon was also prepared from genomic DNA and inserted into the vector using the SacI and BamHI sites.
A DNA fragment containing GFP was inserted into pRS416-stem-loop-HOM2-FLAG-CYC1t using BamHI and XhoI sites (pRS416-HOM2 5 UTR-GFP-CYC1t). Then, a DNA fragment containing the HOM2 3 UTR was amplified by PCR from genomic DNA and inserted into the vector at XhoI site.
A DNA fragment spanning the promoter and the ORF region of HOM2 was amplified by PCR from genomic DNA as a template and inserting this fragment into pRS416-CYC1t URA3 CEN using SacI and XhoI sites. A genome-derived DNA fragment containing the PIG2 3 UTR was then amplified by PCR and inserted into the vector using the XhoI site.
These plasmids were prepared by amplifying a DNA fragment from genomic DNA encoding the region 500 nucleotides upstream of the initiation codon for each gene and inserting each into pRS416- stem-loop-HOM2-FLAG-CYC1t using SacI and SalI sites.
Mortimore, G. E., Lardeux, B. R. & Heydrick, S. J. Mechanism and control of protein and RNA degradation in the rat hepatocyte: two modes of autophagic sequestration. Revis. Biol. Cel. 20, 7996 (1989).
Zaragoza, D., Ghavidel, A., Heitman, J. & Schultz, M. C. Rapamycin induces the G0 program of transcriptional repression in yeast by interfering with the TOR signaling pathway. Mol. Cell Biol. 18, 44634470 (1998).
Kaneko, Y., Toh-e, A. & Oshima, Y. Identification of the genetic locus for the structural gene and a new regulatory gene for the synthesis of repressible alkaline phosphatase in Saccharomyces cerevisiae. Mol. Cell Biol. 2, 127137 (1982).
Nemec, A. A., Howell, L. A., Peterson, A. K., Murray, M. A. & Tomko, R. J. Autophagic clearance of proteasomes in yeast requires the conserved sorting nexin Snx4. J. Biol. Chem. 292, 2146621480 (2017).
Kraft, C., Deplazes, A., Sohrmann, M. & Peter, M. Mature ribosomes are selectively degraded upon starvation by an autophagy pathway requiring the Ubp3p/Bre5p ubiquitin protease. Nat. Cell Biol. 10, 602610 (2008).
Chan, L. Y., Mugler, C. F., Heinrich, S., Vallotton, P. & Weis, K. Non-invasive measurement of mRNA decay reveals translation initiation as the major determinant of mRNA stability. elife 7, e32536 (2018).
Cleary, M. D., Meiering, C. D., Jan, E., Guymon, R. & Boothroyd, J. C. Biosynthetic labeling of RNA with uracil phosphoribosyltransferase allows cell-specific microarray analysis of mRNA synthesis and decay. Nat. Biotechnol. 23, 232237 (2005).
We are grateful to all the members of Ohsumi and Iwasaki laboratories for helpful discussions and critical reading of the manuscript and the Biomaterials Analysis Division, Tokyo Institute of Technology for DNA sequencing. We would like to thank Alexander May for editing the manuscript and Shukun Hotta-Ren for technical assistance; Yuichiro Mishima, Yoshitaka Matsuo and Takashi Fukaya for critical comment on manuscript; Hideki Taguchi for helping polysome analysis; Koji Yoda for providing antibodies. This work was supported in part by a Grant-in-Aid for Young Scientists (B) (JP17K15063 to S.M.), a Grant-in-Aid for Scientific Research (C) (JP15K06949 to T.K.), a Grant-in-Aid for Scientific Research (S) (JP16H06375 to Y.O.), a Grant-in-Aid for Young Scientists (A) (JP17H04998 to S.I.), and a Grant-in-Aid for Challenging Research (Exploratory) (JP19K22406 to S.I.) from the Japan Society for the Promotion of Science (JSPS) and by a Grant-in-Aid for Scientific Research on Innovative Areas Multidisciplinary research on autophagy (JP16H01197 to T.K.), a Grant-in-Aid for Scientific Research on Innovative Areas nascent chain biology (JP17H05679 to S.I.), and a Grant-in-Aid for Transformative Research Areas (B) (JP20H05784 to S.I.) from the Ministry of Education, Culture, Sports, Science and Technology (MEXT). S.M. was also supported by JST, ACT-X (JPMJAX201E). S.I. was also supported by AMED-CREST, AMED (JP20gm1410001), the RIKEN Pioneering Project (Cellular Evolution) and Aging Project, and the Takeda Science Foundation. RNA-Seq analyses was supported by the JSPS Platform for Advanced Genome Science (JP16H06279 to T.K.). DNA libraries were sequenced by the Vincent J. Coates Genomics Sequencing Laboratory at UC Berkeley, supported by NIH S10 OD018174 Instrumentation Grant. Computations were supported by Manabu Ishii, Itoshi Nikaido, the Bioinformatics Analysis Environment Service on RIKEN Cloud, and HOKUSAI Sailing Ship in RIKEN.
Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made. The images or other third party material in this article are included in the articles Creative Commons license, unless indicated otherwise in a credit line to the material. If material is not included in the articles Creative Commons license and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this license, visit http://creativecommons.org/licenses/by/4.0/.
Achieving and maintaining optimal recovery of valuable minerals is not something that happens by chance. Designing a new flotation process, or choosing the right equipment for an existing process, requires experience and a deep understanding of minerals processing and metallurgy. Our solutions are designed in our Engineering and R&D centers by experts specialized in minerals processing and technology development. Using state-of-the-art laboratories and pilot plants for the most demanding test work, our specialists help you evaluate the best options for your process and make informed decisions for the entire life cycle of your plant.
The productivity of your process has a direct effect on the profitability of your operations. Our decades of experience working with metals and minerals processors around the world has enabled us to develop some of the most innovative and productive flotation technologies on the market. Outotec FloatForce technology is a good example: By improving flotation hydrodynamics and pumping performance at high air dispersion rates, FloatForce enhances particle recovery in the flotation cell, while also reducing power consumption and the risk of sanding.
Our extensive service portfolio and customized solutions cover all phases of your flotation plants life cycle. We handle engineering and procurement (EP), engineering, procurement, and supply (EPS) or engineering, procurement, and construction (EPC) projects for flotation plants. Our service scope is always tailored to meet your unique business challenges and to have maximum impact on your performance and productivity. We can help you with everything from increasing productivity, product quality, energy efficiency, and recovery rates to improving water conservation, asset availability, and environmental compliance.
Our flotation cell retrofit uses provenOutotec TankCell technologyand state-of-the-art OutotecFloatForceforced-air technology. In addition to these technologies, the service can also include modern automation solutions as well as tailor-made service solutions.
This is an overview of setting up and conducting a flotation rate test. The test is a means of determining the flotation characteristics of an ore. It is conducted in a laboratory scale cell usually with a volume of two point five litres. The intention is to generate relationships of cumulative recovery, mass pull and grade versus time and use these to evaluate the floatability of metal, mineral and gang. The video is not intended to be a detailed explanation of how to conduct a rate test. For that, download the test procedure shown here below.The purpose of this video is to outline the important aspects of setting up the test to ensure that flotation characteristics are correctly measured and the resulting data can be properly interpreted. How the data is used to determine flotation kinetics is a subject of a separate video. The key aspects are:
This comes standard with 73millimeter diameter rotor and a stator of 80 millimeters inside diameter. The 2.5 L cell has an operating volume of about 2 Land takes 1 kilogram of solids, giving a pulp density of 32% solids by mass. Air is self-induced through a valve on the shaft. Alternately air can be supplied from a compressor and controlled by valve.
What rotor speed is used on the D12 in a flotation test is a matter of personal choice or company standard. The speed affects the degree of agitation and the amount of air going into the cell and the number of air bubbles generated per unit time.
Both mineral and gangue recovery increases with increasing aeration rate. As shown here from an example taken from tests on the Merensky PGM ore from in South Africa.The graphs show platinum group metals as PGMs and gangue recovery with flotation time at six different air rates measured as cubic meters of air per minute per cubic meter of pulp. PGM concentrate mass recovery increase at different rates resulting in the system operating at successively different recovery grade curves.You can see that highest concentrate grade is generated at the lowest air rate but also for the lowest recovery. In other words, lowest aeration rate to generate the best selectivity between mineral and gangue.
These tests were conducted at a fairly high rotor speed of fifteen hundred RPM. If rotor speed is reduced, the same effect is achieved; that is selectivity increases as less air is introduced into the system.However in this case an extra variable is being added due to the smaller degree of agitation.Some companies and operators prefer to run with a lower rotor speed such as nine hundred rpm in an effort to derate the laboratory cell so that its operation is closer to that of a production scale cell.These graphs illustrate that aeration rate changes flotation performance quite significantly.Whatever rotor speed and aeration rate is chosen for consistency make sure that all subsequent tests are performed at the same speed and air rate. The sample can be crushed and milled in the laboratory or it can be obtained from sampling the required stream in the plant.Rate testing a plant grab sample is often known as a hot float.For samples prepared in the metallurgical laboratory, all conditions of grinds, reagents, pH and %solids etc are chosen by the operator.
Hot floats are normally performed as is unless the intention is to add additional reagents or change pulp density and pH.Note that to avoid ageing of the sample and possible oxidation effects, flotation should be done as soon as possible after sample collection. Auto-rotation speed of fifteen hundred RPM, add water until the pulp level is set at fifteen to twenty millimetres below cell overflow lip. The distance will be less at lower rotor speeds. Add and condition all reagents finally adding frother and begin the test by opening the air valve.The aim throughout the test is to maintain the top of the froth bed, level with the overflow lip of the cell. As the test proceeds this is achieved by gradually opening the air valve.When fully opened, froth level is kept at the desired height by raising pub level with make-up water. By keeping the level of the froth consistently equal to that of the overflow lip, the quantity of froth removed is controlled by the paddle design and the number of froth removal sweeps per minute. Details of the number of concentrate collections per minute and timing of the sweeps can be found in EMCs procedure.
The video shows an automatic laboratory flotation cell developed by Outotec at their research centre in Pori in Finland. In this case the cell is 12litres treating 3.5kilograms of ore 25%solids. Rotor speed is 1500 RPM and air is fed at a rate of 5 LPM. The collection paddle is automated and set to collect froth every 10 seconds to a depth of about10 millimetres below the overflow lip.Know the level marks on the side of the cell; the top one being levelled with the overflow lip. Water is added to obtain the required pulp level. The paddle is split so that collection can be from the back of the cell and around the shaft as it moves forward. Paddle shaft and the sides of the cell are washed down with spray water. Level control is automated and both makeup and spray water masses used are recorded throughout the test.
This concludes the overview of the important features of a flotation rate test. View the next video on the meaning and use of kinetics to find out more about or characterization and flotation circuit design.
The flotation machine is one of the most important equipment in the beneficiation process, evergreen company production of the rotor and stator of advanced design, mold design, reasonable chemical stability, impact resistance, tear resistance, light weight, convenient handling, maintenance and replacement, dynamic balance detection stability and enhance the stability of the operation, reduce the partial load, the rotor and the stator is made of high quality rubber, polyurethane, nylon formula three specifications, quality first-class.
The flotation machine is one of the most important equipment in the beneficiation process, evergreen company production of the rotor and stator of advanced design, mold design, reasonable chemical stability, impact resistance, tear resistance, light weight, convenient handling, maintenance and replacement, dynamic balance detection stability and enhance the stability of the operation, reduce the partial load, the rotor and the stator is made of high quality rubber, polyurethane, nylon formula three specifications, quality first-class.
Each model, Eevergreen produce full range of flotation machine stator, impeller. Such as KYF type, XCF type, BFtype, JJF type, BS-K type, GF type, SF type, FLSML type, DTH type, WEMCOtype, DORR type, etc.. The company has the most comprehensive AI Grenada sofar the biggest manufacturer of the world specifications, different types of flotationmachine, to provide comprehensive solutions, all of the stator and impeller can be customized according to the actual conditions of Surveying and mapping, and operating mines tailored rubber, polyurethane formulation, macromolecule nylon formula.
Evergreen Company covers an area of 23000 square meters, the company has 200 employees, including 20 management technical personnel, 20Years experience in mining equipment, Exported to more than 30 countries.
Archimedes principle states that a body immersed in a fluid is subjected to an upwards force equal to the weight of the displaced fluid. This is a first condition of equilibrium. We consider that the above force, called force of buoyancy, is located in the centre of the submerged hull that we call centre of buoyancy. A second condition, known as Stevins law, states that the centre of gravity of the floating body and its centre of buoyancy must lie on the same vertical. For a small angle of inclination the initial and the inclined waterplanes intersect along a line passing through the centroid of the waterplane. For various inclinations the centre of buoyancy travels along a curve whose centre of curvature is called metacentre. For a body floating at the surface, the equilibrium is stable if the metacentre is situated above its centre of gravity.
Archimedes principle states: An object immersed in a fluid experiences a buoyant force that is equal in magnitude to the force of gravity on the displaced fluid. Thus, the objective of underwater vehicle flotation systems is to counteract the negative buoyancy effect of heavier than water materials on the submersible (frame, pressure housings, etc.) with lighter than water materials; A near neutrally buoyant state is the goal. The flotation foam should maintain its form and resistance to water pressure at the anticipated operating depth. The most common underwater vehicle flotation materials encompass two broad categories: Rigid polyurethane foam and syntactic foam.
The term rigid polyurethane foam comprises two polymer types: Polyisocyanurate formulations and polyurethane formulas. There are distinct differences between the two, both in the manner in which they are produced and in their ultimate performance.
Polyisocyanurate foams (or trimer foams) are generally low-density, insulation-grade foams, usually made in large blocks via a continuous extrusion process. These blocks are then put through cutting machines to make sheets and other shapes. ROV manufacturers generally cut, shape, and sand these inexpensive foams, then coat them with either a fibreglass covering or a thick layer of paint to help with abrasion and water intrusion resistance. These resilient foam blocks have been tested to depths of 1000 feet of seawater (fsw) (305m) and have proven to be an inexpensive and effective flotation system for shallow water applications (Figure 10.45).
Polyisocyanurate foams have excellent insulating value, good compressive-strength properties, and temperature resistance up to 300F. They are made in high volumes at densities between 1.8 and 6lb per cubic foot, and are reasonably inexpensive. Their stiff, brittle consistency and their propensity to shed dust (friability) when abraded can serve to identify these foams.
For deep-water applications, syntactic foam has been the foam of choice. Syntactic foam is simply an air/microballoon structure encased within a resin body. The amount of trapped air within the resin structure will determine the density as well as the durability of the foam at deeper depths. The technology, however, is quite costly and is normally saved for the larger deep-diving ROV systems.
Archimede's Principle states that a body immersed in a fluid experiences an upthrust equal to the weight of the fluid displaced, and this is fundamental to the equilibrium of a body floating in still water.
A body floating freely in still water experiences a downward force acting on it due to gravity. If the body has a mass m, this force will be mg and is known as the weight. Since the body is in equilibrium there must be a force of the same magnitude and in the same line of action as the weight but opposing it. Otherwise the body would move. This opposing force is generated by the hydrostatic pressures which act on the body, Figure 3.1. These act normal to the body's surface and can be resolved into vertical and horizontal components. The sum of the vertical components must equal the weight. The horizontal components must cancel out otherwise the body would move sideways. The gravitational force mg can be imagined as concentrated at a point G which is the centre of mass, commonly known as the centre of gravity. Similarly the opposing force can be imagined to be concentrated at a point B.
This is also the weight of the displaced water. It is this vertical force which buoys up the body and it is known as the buoyancy force or simply buoyancy. The point, B, through which it acts is the centroid of volume of the displaced water and is known as the centre of buoyancy.
In other words the mass of the body is equal to the mass of the water displaced by the body. This can be visualized in simple physical terms. Consider the underwater portion of the floating body to be replaced by a weightless membrane filled to the level of the free surface with water of the same density as that in which the body is floating. As far as the water is concerned the membrane need not exist, there is a state of equilibrium and the forces on the skin must balance out.
A density apparatus that uses the Archimedes principle was used to find the density. The hardness was measured by Rockwell hardness using the K scale (steel ball indenter with 3.125 with 1500N load). Resins that were not appropriately cured were found using the Soxhlet extraction apparatus by acetone extraction. Uncured resins were found using the acetone extraction test by using a Soxhlet extraction apparatus. The mass loss due to the ignition of the brake pads was found by considering 510g of the developed sample in a silica crucible heated to 800C and maintained at that temperature for 2h in a muffle furnace. The difference in mass loss was noted and reported. The pad swell on heating was found by considering a sample of size of 10mm10mm4mm that was maintained in a hot air oven for 40min at 2003C. The brake pad samples were cut to a 50mm25mm size and immersed in water in ambient conditions for 30min. For both tests, the difference in thickness was noted. The above tests were performed as per IS2742 Part 3. The hot shear strength was measured by heating the pad to 200C and maintaining at that temperature for 30min while cold shear strength was found by testing the brake pads under ambient conditions. The above shear tests were performed as per ISO 6312. The image analysis technique used by industries was followed to find the adhesion. Porosity was found by placing brake pad samples of 625mm2 in a desiccator for 24h and then soaking it in a preheated SAE 90-grade oil bath at 90C1C for 8h. After the required time, the heater was switched off, and the cooling of the samples was performed at room temperature. It was carried out according to JIS D 4418 . To check the consistency, three samples were tested and the average values reported. A quick assessment of the tribological characteristics was done using the Chase test following IS 2742 Part 4 for a brake pad sample size of 625mm2 that was made to slide against a 280mm diameter cast iron drum [20, 21].
The fade, recovery, and effectiveness characteristics of the developed brake pads were analyzed using a full-scale inertia brake dynamometer as per the JASO C-406 schedule. The parameters were an inertia of 50kgms2, a rolling radius of 0.293m, an effective radius of 0.103m, a pad thickness with a backplate of 15.1mm, and a mating surface of cast iron. The calibration of the data acquisition system used to record the data was done through process variable equipment, namely pressure sensors, temperature sensors, speed sensors (encoders), and torque measurement (load cell) as per the National Traceability Standards of NABL-certified laboratories. For consistency in results, two samples were tested. If the error was more than 8%, then a third sample was tested. The establishment of 90% contact between the pad and disc was done by a bedding test at a 50km/h speed with 3.0m/s2 deceleration with 200 applications having an initial brake temperature of 100C. Three effectiveness studies were performed. In the first effectiveness study, two different speeds (50 and 100km/h) with two different decelerations (3.0 and 6.0m/s2) with an initial temperature of 80C were considered. In the second effectiveness study, three different speeds (50, 100, and 130km/h) with two different decelerations (3.0 and 6.0m/s2) with an initial temperature of less than 80C were considered. Finally, in the third effectiveness study, three different speeds (50, 100, and 130km/h) with two different decelerations (3.0 and 6.0m/s2) with an initial temperature of less than 100C were considered. For all effectiveness studies, the number of applications was repeated until the measurements for four or more points were equal within the deceleration range. For the above test, the air blower was kept off.
Two decelerations (3.0 and 6.0m/s2) were carried out at a 100km/h vehicle speed for the fade test with a braking interval of 35s for 15 applications in each deceleration with an initial brake temperature of 80C under blower off conditions. In the case of recovery, two decelerations (3.0 and 6.0m/s2) were repeated at 120s intervals for a 50km/h speed, with the air blower on and with an initial brake temperature of less than 80C. The was measured and the average values reported. The fade and recovery rates were calculated as per the formulas that are given in Eqs. (1), (2). The pad and rotor wear were reported based on the after test measurement using a digital micrometer [18, 19].
Where rmax and rmin are the maximum and minimum coefficients of friction measured during the recovery cycle, and fmax and fmin are the maximum and minimum coefficients of friction measured during the fade cycle.
The K-type (button type) thermocouple was used to measure temperature. For measurement, drilling was performed through the spigot hole and a thermocouple was placed just beneath the pad surface. The final and initial temperatures were noted. Another through-hole was made for contact with the disc to measure the braking interface temperature. In the current study, the pad temperature was measured and reported. In some instances, as prescribed by the customer or standard, a temperature rise in the disc would be reported. A scanning electron microscope (Tescan VEGA 3LMU SEM) was used to analyze the worn surfaces of the tribological-tested brake pad samples.
Now let us attempt another application using Archimedes' principle. Suppose we have an open-ended tube fixed to a vertical wall and extending horizontally into a liquid, and the central longitudinal axis of the tube is at some depth, h, as in Figure 2-5. What is the longitudinal axial load in this horizontal tube (neglecting bending loads due to gravity for now)? We can see that a pressure load on the end of the tube is acting on the cross-sectional area, so we know there is a compressive axial load in the tube. Can we use Archimedes' principle to determine that load? No, we cannot, because Archimedes' principle says nothing about horizontal hydrostatic loads. Also we cannot simply multiply the pressure by the cross-sectional area, as we did previously, because it is apparent from the figure that the pressure on the cross section is not the same at all points since the pressure varies with depth. So, let us see how we calculate the load on the end due to the liquid pressure, which varies with depth.
We could show this more easily if it were a solid bar with a rectangular cross section, but since our interest is in tubes, we might as well see the details of how it is done. Since pressure varies with depth, we can express the pressure at some point on the tube end as follows:
Note carefully the orientation of our coordinate system, because we have adopted the convenient system mentioned earlier for our use in well-bore calculations, and it appears to be upside down to what we are accustomed to seeing; that is, the z-axis is positive downward. The angle, , is measured counterclockwise from the positive z-axis. Since the pressure varies over the area of the tube, the force due to the pressure on the end of the tube is the pressure integrated over the area of the tube:
From this result, we see that the force on the end of the tube is equal to the pressure at the center of the tube times the cross-sectional area of the tube. Is this a general result for any tube, or is it specific for a horizontal tube face? This can be generalized to any inclination and is an important result in fluid statics, in that the force of a fluid of constant density on a submerged flat surface is equal to the pressure at the centroid of the surface times the area of the surface.
Particle volumes can be determined directly by Archimedes' principle, which compares the weight of a particle measured in water to that measured in air (Hughes, 2005). This technique is only practical for lapilli-sized particles, however, and is best suited to nonporous material. Another method for measuring lapilli-sized particles is gas (He) pycnometry, which is usually used for density measurements and is based on gas displacement principle (eg, Klug etal., 2002). For uncoated porous particles, both Archimedes-based methods and gas pycnometers measure the skeletal volume, which is the particle volume without considering isolated internal vesicles. More advanced methods for direct volume measurements include 3-D laser and CT scanners, which can provide very detailed information about individual particle size and shape. Laser scanners can reconstruct the external envelope of a particle surface in 3-D at a given resolution (eg, 400 points per square inch with resolution of 100m for the NextEngine 3D Scanner). CT scanners use X-rays to create shadow projections of the particle on an X-ray sensitive camera and to reconstruct particle 2-D CT slices and 3-D model, which also contains information about the internal structure of the particle (Bagheri etal., 2015). The main disadvantages are that 3-D laser and CT scanners are not widely accessible, have resolution limits, and require considerable pre- and postprocessing time to acquire a 3-D model of the particle (Bagheri etal., 2015).
Despite recent major technical breakthroughs related to rapid acquisition of 3-D models of irregular particles by CT scanning (eg, Garboczi etal., 2012), volume measurements based on direct methods are still not practical, and therefore, indirect methods are used to estimate particle volume. Particle volume, V, can be simply calculated using the diameter of the volume-equivalent sphere deq:
One method for estimating deq is to average the tridimensional length of the particle known as particle form dimensions (Bagheri etal., 2015 and references therein). Form dimensions are defined and noted as L: longest, I: intermediate, and S: shortest length of the particle (Fig.1A). These dimensions can be measured with a caliper (or, in general, the distance between two parallel plates tangent to the particle edges) and are called Feret lengths (diameters). Several protocols exist for measuring form dimensions of irregular particles. The most commonly accepted is the standard protocol proposed by Krumbein (1941). First, the longest dimension of the particle, L, is measured; the longest dimension perpendicular to that is I, and, finally, S is the longest dimension perpendicular to both I and L. In contrast, Blott and Pye (2008) defined L, I, and S with respect to the longest, intermediate, and shortest edge dimensions of the minimum bounding box enclosing the particle (Fig.1B). The accuracy of these protocols is, however, highly dependent on the ability of the operator to identify the perpendicular directions along which L, I, and S should be measured. Recently, Bagheri etal. (2015) introduced a projection area protocol, which, unlike others, does not require form dimensions to be perpendicular to each other. Instead, form dimensions are measured particle projections that have maximum and minimum areas (Fig.1C). L and I are defined as the largest and smallest dimensions measured on the maximum area projection, and S corresponds to the smallest dimension measured in the minimum area projection. Form dimensions measured using the projected area protocol have low operator-dependent errors and provide accurate estimates of particle volume and surface area (Bagheri etal., 2015).
Figure1. Schematic illustration of different protocols used to measure form dimensions (L, I, and S) of an irregular volcanic particle. (A) The standard protocol proposed by Krumbein (1941). (B) The minimum bounding box protocol of Blott and Pye (2008). (C) The projection area protocol proposed by Bagheri etal. (2015). Top and bottom projections are the maximum and minimum area projections, respectively.
It should be noted that using geometric averaging for estimating deq is equivalent to approximating particle shape with its dimension-equivalent ellipsoid. The dimension-equivalent ellipsoid is here considered as an ellipsoid with similar form dimensions (flatness and elongation ratios) as the particle. The accuracy of Eqs.  and  has been benchmarked by measuring the volume of coarse ash- and lapilli-sized dense and vesicular volcanic particles with scanning electron microscope (SEM) micro-CT and 3-D laser scanning techniques (Bagheri etal., 2015). Both Eqs.  and  overestimated deq for all considered particles compared to SEM micro-CT and 3-D laser scanner measurements. The overestimation of the arithmetic mean was on average 16% (max. 60%), whereas the geometric mean performed better with an average overestimation of 12% (max. 50%).
Another common method for estimating deq uses measurements obtained from particle image analysis (Leibrandt and LePennec, 2015). In fact, fully automated particle sizers can now produce and analyze 100s of particle images within a few minutes. The images are particle projections (silhouettes), from which several 2-D size characteristics can be measured, including projection area, AP, projection perimeter, P, and circle equivalent diameter d2D (Fig.2).
Figure2. Commonly measured variables through image analysis. Aprojection area, Pprojection perimeter, d2Dcircle equivalent diameter, lminminimum caliper length, lmaxmaximum caliper length, Didiameter of the largest inscribed circle, and Dcdiameter of the smallest circumscribed circle.
Circle equivalent diameter (also known as Heywood diameter), d2D, is the diameter of a circle with the same area as the particle projection and is commonly used to approximate deq. However, d2D is strongly dependent on the particle orientation in the captured projection. As a result, obtaining a better approximation of deq by d2D requires a large number of particle projections in different orientations. Needless to say, this is time-consuming and in some cases, impossible, due to difficulty of manually handling fine ash particles; for this reason, most studies use only a single particle projection to estimate volume. Under these conditions, d2D can be on average 26% (max. 80%) higher than the actual deq of the particle (Bagheri etal., 2015). This overestimation can be decreased if more particle projections are considered; for example, by using 1000 projections, the average error can be reduced to 12% (max. 40%).
The densities of the prepared glass nanocomposite samples were measured by Archimedes principle using acetone as an immersion liquid. Molar volume of a substance is the volume of one mole of that substance. Molar volume of a substance is defined as the ratio of its molecular or atomic weight, whichever is suitable to its density. The relationship between density and composition of an oxide glass nanocomposite system can be expressed in terms of an apparent molar volume of oxygen (VM) for the system, which can be obtained using the formula
The dispersed phase equation (T2.5), was constructed to satisfy Archimedes principle. This principle is immediately satisfied when considering the case of statics with no solids stress. For example, consider a dispersed gasliquid system, with bubbles of density s and a fluid of density g. Then clearly the force per unit length acting on the bubbles is
The relation between B and A can be obtained by comparing Eq. (2.9) to Eq. (2.26) for the case of zero velocity gradient, zero gravity, and zero wall friction. The difference between these equations is the porosity that multiplies the gradient of pressure in the model in which the pressure drop is taken to be in the gas and solid phases. As an alternative, B can be obtained directly from the Ergun Equation (2.10). A similar derivation for Model C shows that the friction coefficients for Models B and C are equal.
Note that there is a lack of symmetry between the dispersed phase and the continuous phase momentum balances. Therefore, in applying these equations to describe boiling going from all liquid to all vapor, it is necessary to switch from liquid continuous phase to gas at some appropriate volume fraction of steam. Equations (T2.5) and (T2.6) can also be used to model flow of neutrally buoyant particles.
The effective axial load as we define it, is calculated using Archimedes' principle, in that the buoyant force is equal to the weight of a fluid displaced by the submerged portion of a body. For convenience we use a buoyancy factor, kb, based on the density difference between that of the body and the fluid. The buoyancy factor multiplied by the weight of the casing in air gives the buoyed weight of the casing.
As already stated, the effective axial load is a fictitious load except for a single point in a casing string, the very top of the string. Why does anyone use it then? Most likely because it is so simple, but more disturbingly is the possibility they do not understand what it is. As to its simplicity, yes, but consider a case where the fluid inside the casing is different from the fluid in the annulus. How does that affect our simple buoyancy factor equation above? Clearly stated, the effective axial load is of no use in determining the axial load in casing design because it does not give us the axial load! That said, does the effective axial load have any use at all? The answer to that is emphatically, yes. It is used in determining the point of neutral stability for lateral buckling in tubular strings. It has been used correctly for many years in calculating the length of drill collars needed to prevent lateral buckling in drill pipe. We will discuss lateral buckling in Chapter 6, and again, see Appendix A for more detailed discussion on buoyancy.
Doesanyone have experience with adjustment of rotor engagementfor Wemco rougher flotation cells? Currently our rotor engagement is varying between 115and 145 mm and I am not sure which is correct or optimal to provide enough air flow to maximize the recovery. (rotor engagement is the vertical distance that the rotor is encased within the draft tube)
All cell is in same row engagements were different and varying from 115 to 145 mm. We're planning to measure the air flow of each cell using a hand anemometer and see any correlation with engagement. The most difficult thing is to correlate engagement and recovery as all cells conditions and performance are different. Some cells were sanded heavily specially underneath of false floor. It takes time to clean up these build ups and descale them clean. We partially get to them on long shutdowns as operations wants them to hold production to maximum.
How often do you think impeller needs to be changed? Our agitators are now 4 years old (new at startup) and no significant wear is so far observed for rotors. Do them not wear out fast because our engagement is not right or it it just like that?
I think you should study the gasdispersion en the rougher flotation cells. Results of this kind of studies indicate that air flow rates, air flow numbers and air flow velocities may be differentfrom cell to cell but they can be inthe range of a typical flotation cell. It is well known that gas dispersion is a key aspect in the performance of a flotation cell.
It is important to mention that the air flow rate can be calculated from the air velocity measured in a pipe placed intheair inlet of the flotation cell. In some cases it could be useful to employ two different kinds of anemometers depending on the flotation cell. For instance, a vane anemometer can be consideredto determine theair velocity for induced-air cells. A hot wireanemometer is an interesting option to evaluate the air inlet inforced-air flotation cells. The evaluation should include information on the superficial gas velocity because they can provide information of air dispersion in a flotation cell.
You mentioned a sanding problem in the flotation cells. The problems can be associated to the problems in the gas dispersion, presence of coarse particles, variations in the slurry density and inappropriate operation the flotation cells. Try to get information on the parameters mentioned above. If there is not enough data, it is important to take samples to get the information.
There's been work by various groups showing improved performance by a range of control actions from supercharging some types of Wemco cells (Codelco) to throttling. Remember the impact of engagement on pumping action impacts more than just aeration. Issues with engagement could be causing the sanding because of reduced pumping action. Do you have an engagement recommended the manufacturer?
If I was to make a recommendation when you perform your air flow test it would be to limit the factors that would affect air flow such as draft tube cleanliness, bottom side sanding, even mechanical problems that are not so obvious such as belt slippage, and disperser skirt condition. Once all the variables are removed then you can determine if engagement is part of your problem or if perhaps it is something else.
The Wemco unit is a naturally-aspirated mechanical flotation machine. Air is induced via a standpipe and an inlet pipe open to the atmosphere. For Wemco flotation machines of constant rotor diameter, hydrodynamic parameters are defined as the relationship between air ingestion and primary liquid circulation as functions of rotor submergence, rotor engagement, and rotor speed. In turn, rotor submergence is defined as the vertical distance from the top of the rotor to the froth- pulp interface, and rotor engagement as the vertical distance that the rotor is encased within the draft tube.
At constant rotor submergence, an increase in rotor speed results in an increase in aeration (or superficial gas velocity), and specific power. At constant submergence and rotor speed, liquid circulation varies almost linearly with rotor engagement. At constant rotor speed, an increase in rotor submergence results in a decrease in air induction, with a simultaneous increase in the draft tube liquid circulation (the latter being synonymous with absorbed power and often manifested as specific power input). The rotor-disperser combination serves the same purpose as a rotor-stator arrangement in a forced-air machine.
I have been using a variety of Wemco cells (ranging from 40 cu.ft. covered to 3500 cu.ft. tank versions) at a number of operations for over 20 years and can tell you the challenges you are facing are not unique. The technical explanations and advice offered above are valid, so I will provide my operational experience:
Your best weapon against poor performance is training - teach your operators and mechanics how the cells work, why it is important to wash them out occasionally, and what to look for when they are inspecting the cells (they have to climb inside the cell to properly inspect). If your cells are tiny and/or covered, invest in a flexible "snake" camera for visual inspection. It does not take much neglect to trigger a performance decline.
DISCLAIMER: Material presented on the 911METALLURGIST.COM FORUMS is intended for information purposes only and does not constitute advice. The 911METALLURGIST.COM and 911METALLURGY CORP tries to provide content that is true and accurate as of the date of writing; however, we give no assurance or warranty regarding the accuracy, timeliness, or applicability of any of the contents. Visitors to the 911METALLURGIST.COM website should not act upon the websites content or information without first seeking appropriate professional advice. 911METALLURGY CORP accepts no responsibility for and excludes all liability in connection with browsing this website, use of information or downloading any materials from it, including but not limited to any liability for errors, inaccuracies, omissions, or misleading statements. The information at this website might include opinions or views which, unless expressly stated otherwise, are not necessarily those of the 911METALLURGIST.COM or 911METALLURGY CORP or any associated company or any person in relation to whom they would have any liability or responsibility.
Flotation is about creating the proper energy dissipation rate in the cells to obtain optimal contact between the air bubbles and the particles for extracting the minerals. The function of the rotor/stator is to make bubbles from the forced air, suspend the particles, and create an environment for bubbles and particles to make contact and rise to the top as froth for concentration and collection.
Our forced-air flotation design features a streamlined, high-efficiency rotor that works as a very powerful pump. Working together the stator, these components generate an energy-intensive turbulence zone in the bottom of the cell. The forced-air design allows for control of the air flow. The well-defined turbulence zone results in multiple passes of unattached particles through the highest energy dissipation area of the cell where fine particles are driven into contact with the air bubbles.
The stator design, in addition to providing good separation of the cell zones, also serves to redirect the rotor jet uniformly across the tank. This allows dispersion, or distribution, of the maximum amount of air into the cell without disturbing the surface an important consideration for fine particle recovery. The air dispersion capabilities of our Dorr-Oliver cell design exceed all competitive forced-air designs.
By containing the intense circulation energy at the bottom of the cell, the upper zones of the cell remain quiescent, or passive, to maximise recovery of marginally attached coarse particles and minimise the carriage of undesired material.
We have equipped our forced-air flotation tank cells with a uniquely designed, high-efficiency radial launder system that accelerates froth removal as it reaches the surface. Bubble-particle aggregates travel vertically through the froth lattice. The high-efficiency radial launder is shaped to receive the froth uniformly from the cell surface, as well as from the typically heavy-loaded area near the centre of a forced-air machine. On passing over the lip, the froth accelerates to the perimeter of the cell. This unique design rarely requires launder water.
The two factors having the strongest impact on a flotation circuits performance are metallurgical recovery and flotation cell availability. Our forced-air flotation machines provide superior performance in both of these important areas, while offering additional, distinct advantages.
Superior metallurgical performance: Intense recirculation in a well-defined mixing zone multiplies the chances of contact between mineral particles and air bubbles, providing for greater mineral recoveries and higher concentrate grades.
Greater availability: Non-clogging design of the rotor reduces maintenance requirements, minimising failure, and increases availability. Our flotation mechanisms also can be removed for maintenance without process interruption.
Low reagent costs: Air is a natural reagent in the flotation process. Having a wide air dispersion capability permits you to fine-tune your flotation plant to deliver the optimum value for your process.
FLSmidth provides sustainable productivity to the global mining and cement industries. We deliver market-leading engineering, equipment and service solutions that enable our customers to improve performance, drive down costs and reduce environmental impact. Our operations span the globe and we are close to 10,200 employees, present in more than 60 countries. In 2020, FLSmidth generated revenue of DKK 16.4 billion. MissionZero is our sustainability ambition towards zero emissions in mining and cement by 2030.Get in Touch with Mechanic