Measuring mRNA Levels Over Time During the Yeast S. cerevisiae Hypoxic Response

Summary

Here, we present a protocol using RNA-seq to monitor mRNA levels over time during the hypoxic response of S. cerevisiae cells. This method can be adapted to analyze gene expression during any cellular response.

Abstract

Complex changes in gene expression typically mediate a large portion of a cellular response. Each gene may change expression with unique kinetics as the gene is regulated by the particular timing of one of many stimuli, signaling pathways or secondary effects. In order to capture the entire gene expression response to hypoxia in the yeast S. cerevisiae, RNA-seq analysis was used to monitor the mRNA levels of all genes at specific times after exposure to hypoxia. Hypoxia was established by growing cells in ~100% N₂ gas. Importantly, unlike other hypoxic studies, ergosterol and unsaturated fatty acids were not added to the media because these metabolites affect gene expression. Time points were chosen in the range of 0 - 4 h after hypoxia because that period captures the major changes in gene expression. At each time point, mid-log hypoxic cells were quickly filtered and frozen, limiting exposure to O₂ and concomitant changes in gene expression. Total RNA was extracted from cells and used to enrich for mRNA, which was then converted to cDNA. From this cDNA, multiplex libraries were created and eight or more samples were sequenced in one lane of a next-generation sequencer. A post-sequencing pipeline is described, which includes quality base trimming, read mapping and determining the number of reads per gene. DESeq2 within the R statistical environment was used to identify genes that change significantly at any one of the hypoxic time points. Analysis of three biological replicates revealed high reproducibility, genes of differing kinetics and a large number of expected O₂-regulated genes. These methods can be used to study how the cells of various organisms respond to hypoxia over time and adapted to study gene expression during other cellular responses.

Introduction

Many organisms respond to hypoxia, or low O₂, by altering gene expression ^1,2,3. This response helps cells cope with the lack of a substrate critical for aerobic respiration and for several biosynthetic reactions, but also with a changing redox state ⁴. Several microarray studies performed in S. cerevisiae show that the mRNA levels of hundreds of genes change in response to hypoxia ^{5,6,7,8,9,10,11,12}. Recently, RNA-seq was used to characterize gene expression changes over time during hypoxia ¹³. Here, the experimental details are presented and discussed.

Hypoxia can be achieved in various ways, each producing a different level of O₂. Here, hypoxia was established by continuously flowing ultra-high-purity N₂ into flasks, which lowers [O₂] dissolved immediately with reproducible kinetics ¹⁰. It is possible that there are some O₂ molecules present that contribute to metabolism and gene expression but this environment is considered very close to anaerobic. In the absence of O₂, yeast cells cannot biosynthesize heme, ergosterol and unsaturated fatty acids ^4,12,14. Thus, previous studies have included these metabolites when growing yeast without oxygen ^5,10,15. However, many hypoxic responses are mediated by depletion of these metabolites and thus replenishing them reverses the hypoxic gene expression responses ^12,16. In order to mimic natural hypoxia, these metabolites were not added to the media. In the short time that cells were exposed to hypoxia without the presence of these essential metabolites, there was no noticeable increase in cell death (data not shown) nor a prolonged stress response ¹³.

The response is also dependent upon the strain and its genotype. Especially important are the alleles of the known regulators of hypoxic responses ². The S288C strain background is highly desired so that results can be compared to the other genomic studies performed with this strain. However, S288C contains a partial loss-of-function allele of the HAP1 gene ¹⁷, a transcriptional regulator critical for the hypoxic response. This allele was repaired in S288C using a wildtype copy from the Σ1278b strain background ¹¹.

Gene expression is highly dependent upon the cellular environment. Thus, when performing genome-wide mRNA analysis, it is important to maintain a constant environment while varying another parameter such as time, stimulus, or genotype. To achieve highly reproducible results, consider these three practices for the study and all of its biological or technical replicates. First, the same experimenter(s) should carry out the study, since technical practices may vary across experimenters. Second, the same batch of ingredients should be used in the growth media as each batch has a slightly different composition that can affect gene expression. Third, to minimize cell cycle effects, each time point should consist of asynchronous cells in the mid-log phase of growth (1 - 2 x 10⁷ cells/mL).

When characterizing a complex response like the gene expression response to hypoxia, a time course is advantageous for determining the kinetics of various events. Specific time points should be chosen that will capture the major changes of the response. In this study, time points between 0 and 4 h were observed, because past experiments revealed widespread changes in gene expression during this period ¹³.

To measure global gene expression, RNA-seq was used ^18,19. This method uses next-generation sequencing to determine the relative abundance of each gene's transcript. Compared to DNA microarray analysis, RNA-seq exhibits higher sensitivity (to detect less abundant transcripts), a greater dynamic range (to measure greater fold changes) and superior reproducibility (to accurately follow gene expression over time). Typically, most cellular RNA is ribosomal RNA so many methods have been developed to enrich for specific RNA species ²⁰. Here, poly-T beads were used to purify poly-A-containing mRNA transcripts, though the various commercially- available rRNA depletion kits could also be effective in mRNA enrichment.

Here, the S. cerevisiae gene expression response to hypoxia was characterized. Cells were exposed to hypoxia and then sampled at eight time points (0, 5, 10, 30, 60, 120, 180 and 240 min). To confirm reproducibility and to identify statistically changed transcripts, three biological replicates were performed. RNA was extracted by mechanical disruption and column purification, and then processed for RNA-seq analysis. The post-sequencing pipeline is described and programming scripts are provided that allow exact replication of the analyses performed. Specifically, Trimmomatic ²¹, TopHat2 ²², HTseq ²³, the R statistical environment ²⁴, and the DESeq2 package ²⁵ were used to process the RNA-seq data and to identify 607 genes that change significantly during hypoxia. Principal component analysis (PCA) and gene expression of the replicates indicated the reproducibility of the technique. Clustering and heatmaps revealed wide-ranging expression kinetics, while gene ontology (GO) analysis showed that many cellular processes, like aerobic respiration, are enriched in the set of oxygen-regulated genes.

Protocol

1. Inducing Hypoxia

1. One day or more before the hypoxia time course: Prepare the incubator, cell filtering system, vacuum, gas tank, flasks, stoppers, glass tubing, and tubing, as in the Materials Table.
2. Place the N₂ tank, incubator, vacuum and filtering system in close proximity, to enable quick processing of cells.
3. Prepare sterile liquid YPD media (1% Yeast Extract, 2% peptone, 2% glucose) by mixing the components in a glass bottle and autoclaving.
4. Plan layout of flasks in the incubator.
   NOTE: From the N₂ tank, the first flask will be a water trap, the second flask will be the last time point, the third flask will be the second-to-last time point and so on until the last flask which is a final water trap. Figure 1 shows the order and connections between the flasks.
5. The day before the time course, inoculate a yeast colony into a culture of 5 mL liquid YPD within a sterile test tube. Specifically, pick up a full colony from a plate using a sterile applicator stick. Place the stick into the liquid YPD and shake the stick until most of the cells are in suspension.
6. Rotate the tubes at 30 °C overnight (~16 h).
   NOTE: After 16 h, wildtype cells will achieve saturation (~2 x 10⁸ cells/mL), indicated by a very cloudy culture. Other strains may take longer to reach saturation and should be tested by measuring cell concentration with a spectrophotometer, as indicated in Step 1.9.4.
7. On the day of the time course, after 16 h of incubation, dilute the saturated culture 1:50 by using a 5 mL pipet to place 4 mL of overnight culture into 196 mL of liquid YPD within a sterile 500 mL flask. Grow with shaking at ~200 rpm for 4 h at 30 °C.
   NOTE: After 4 h, a wildtype culture will reach a mid-log concentration of ~1-2 x 10⁷ cells/mL.
8. During the 4 h incubation, label the flasks (for hypoxia and water traps) and the 50 mL collection tubes. If the incubator in step 1.7 above is not used for the hypoxia time course, turn on the other incubator and set to 30 °C.
9. Just before subjecting cells to hypoxia:
   1. Add ~50 mL water to two of the sterile 250 mL flasks which will serve as the water traps.
   2. Fill ¹/₃ of an ice bucket with liquid nitrogen and submerge a polystyrene foam rack to hold 50 mL centrifuge tubes.
   3. Set up the filter system for collecting cells. Add a sterile filter disc onto the bottom unit of the filtration system using sterile tweezers, being careful to only touch the edges of the filter disc. Place the filter top unit onto the filter bottom unit and secure with the clamps provided with the system. Ensure that the bottom and top units are aligned so that there is a tight seal and no leakage.
   4. Measure the cell concentration with a spectrophotometer. Dilute cells so that they can be measured in the linear range of the spectrophotometer – OD₆₀₀ between ~0.2 - 0.6, depending on the spectrophotometer.
      NOTE: One OD₆₀₀ unit is ~3 x 10⁷ cells/mL, depending on cell morphology and spectrophotometer. A concentration of ~1-2 x 10⁷ cells/mL is expected, corresponding to OD₆₀₀ of ~0.33 - 0.66.
   5. Quickly obtain the time 0 sample.
      1. Connect the filter system to a strong (~10 mbar) vacuum via vacuum tubing and a 1,000 mL flask trap. Turn on the vacuum. Pour the culture (usually ~20 mL) into the top unit and wait for liquid to be pulled through the filter (~10 s, depending on the strength of the vacuum).
      2. Carefully remove the filter disc with clean tweezers and place into a 50 mL centrifuge tube. Immediately place the centrifuge tube into the rack submerged in liquid nitrogen. Importantly, do not pre-chill the tubes before inserting the filter as the tubes will explode.
      3. After >30 s, place the tube into a -80 °C freezer for later RNA extraction. After each filtration, clean and reassemble the filter system. Rinse the system by pulling water through for 10 s without a filter disc. Finally, wipe dry the system with a paper towel.
   6. Dilute the 4 h culture into different flasks for the various time points, so that each flask reaches mid-log concentration (~1-2 x 10⁷ cells/mL) at the indicated time point of hypoxia.
      NOTE: Table 1 shows dilutions that were used for the wild-type S288C HAP1⁺ haploid strain. It is recommended to test each strain in these hypoxic culture conditions and adjust the dilutions accordingly.
   7. Once cells and media are added to flasks, replace the aluminum foil covering the flask opening with a stopper containing two glass tubes inserted.
   8. Place the flasks in the incubator in the layout determined earlier.
   9. Securely connect the tubing as shown in Figure 1.
   10. Check that all the stoppers are tightly pushed into the flask openings.
10. At time 0, open the regulator valve. Then set the flow meter to 3 L/min.
    NOTE: Bubbling will be observed in both water flasks, indicating proper gas flow. If this is not the case, check that all stoppers are tight and tubing is properly attached.
11. Close the incubator and set the shaking speed to ~200 rpm. Start a timer.
12. Before each time point, set up the filter system as describe above in step 1.9.3.
13. At each time point (e.g., 5 min, 10 min, etc.), have two people quickly process the cells as follows:
    1. First person: remove the appropriate flask from the holder and take out the stopper (with glass tubing attached).
       NOTE: This will not break the flow of N₂ to any of the remaining culture flasks but will temporarily break the flow to the last water trap.
    2. Second person: Pipet 1 mL of culture and dispense into a cuvette. Reconnect the tubing from the culture flask that is now at the end of the line to the last water trap (One tube and one stopper will be removed in the process.). Then, measure cell concentration in the cuvette, as described above in step 1.9.4.
    3. First person: Pour the remainder of the culture into the vacuum filtration system, and then perform filtering and freezing as described above in step 1.9.5.
14. When finished with all the time points, turn off the N₂ gas. First close the regulator and wait for the pressure to release, and then turn off the flow meter. Finally, disassemble the system and clean all materials with water and then 70% ethanol.

2. RNA Extraction

1. Prepare RLT buffer for RNA column purification, as described in the Materials Table.
2. Prepare a working solution of DNase by adding 10 µL of the DNase stock solution to 70 µL of buffer RDD per sample (see Materials Table).
3. Remove the 50 mL tubes containing filters and cells from the -80 °C freezer and place on ice to thaw (~15 min). Perform the following steps with tubes on ice.
4. Label 2 mL screw-cap tubes and place them on ice, one tube for each sample. Add ~0.6 mL of acid-washed beads to each tube, using a 1.5 mL microcentrifuge tube to scoop and measure the beads.
5. Add 0.6 mL of cold RLT buffer to the 50 mL tubes.
6. Pipet up and down to remove cells from filter and suspend into solution. Avoid letting the filter remain in the solution as the filter will soak up the solution. Remove all solution absorbed into the filter by using the pipet tip to squeeze the filter against the wall of the tube.
7. Transfer all of the liquid from the 50 mL tube to the screw-cap tubes containing beads.
8. Place the screw-cap tubes into a bead mill homogenizer and run for one min.
   Immediately place the tubes on ice for three min. Again, place the tubes into the homogenizer and run for one min.
9. Remove the tubes from the homogenizer and place on ice for 5 min. The beads will settle to the bottom of the tubes.
10. Perform the remainder of the steps at room temperature.
11. With a pipet, transfer only the lysate (~350 µL), avoiding the beads, to a new 1.5 mL microcentrifuge tube.
12. Microcentrifuge for 2 min at max speed and then transfer the supernatant to a new 1.5 mL microcentrifuge tube.
13. Add 1 volume of 70% ethanol (made from 200 proof molecular biology-grade ethanol). Mix well by pipetting.
14. Transfer the solution and any precipitate to an RNA column that has been placed inside a 2 mL collection tube.
15. Centrifuge for 15 s at ≥12,000 x g and discard the flow-through (the liquid in the tube).
16. Add 350 µL of buffer RW1 to the column. Centrifuge for 15 s at ≥12,000 x g and discard the flow-through.
17. Add 80 µL DNase I incubation mix to the column membrane (do not get on sides of tube) and allow to sit for 15 min.
18. Add 350 µL of buffer RW1 to column. Microcentrifuge for 15 s at ≥12,000 x g and discard the flow-through.
19. Add 500 µL of buffer RPE to the column. Microcentrifuge for 15 s at ≥12,000 x g and discard flow-through.
20. Add 500 µL of buffer RPE to the column. Microcentrifuge for 2 min at ≥12,000 x g.
21. Carefully remove the column and place into a new 2 mL collection tube. Microcentrifuge at full speed for 1 min (to completely dry the membrane).
22. Discard the collection tube and place the column into a 1.5 mL microcentrifuge tube. Add 30 µL of RNase-free water to the column membrane.
23. Elute the RNA from the column by microcentrifuging for 1 min at ≥12,000 x g. When loading the columns into the microcentrifuge, make sure the lids of the collection tube are facing in the direction the centrifuge spins, to prevent the lids breaking off.
24. Add another 30 µL of RNase-free water to the column membrane, while keeping the column in the same microcentrifuge tube.
25. Microcentrifuge for 1 min at ≥12,000 x g, so that the final eluate volume is 60 µL.
26. Store each 1.5 mL tube containing 60 µL of RNA in the -80 °C freezer.

3. Determining RNA Concentration and Quality

1. Measure the concentration of RNA and DNA in each RNA sample, using (a) a fluorometer, (b) DNA- and RNA-specific fluorescent dyes and (c) assay tubes, as listed in the Materials Table. Follow the instructions of the fluorometer and the dyes.
2. Alternatively, measure the nucleic acid concentration with a UV spectrophotometer set at 260 nm.
   NOTE: An A₂₆₀ reading of 1.0 is equivalent to ~40 µg/mL single-stranded RNA.
3. Test RNA quality by running the RNA on a commercial nucleic acid analyzer (see Materials Table) or on a standard formaldehyde/agarose gel ²⁶.

4. RNA-seq Analysis

1. From the total RNA stock, make up 21 µL of 100 ng/µL RNA in RNase-free water. Use 1 µL of this 21 µL to test the RNA concentration as above. Submit the remaining 20 µL (2 µg) total RNA to an external sequencing center for mRNA enrichment, strand-specific library preparation (with eight or more barcodes for multiplexing) and sequencing (see Materials Table). Alternatively, locally perform mRNA enrichment and library preparation, and then submit the library for sequencing.
2. Pool eight or more multiplexed samples and sequence in one lane of a next-generation sequencer.
3. Download the FASTQ file containing all of the sequence reads and the corresponding index reads. Also, download the text file containing the multiplex barcodes.
   NOTE: The index reads may be present in a separate FASTQ file. Place all of these files into one directory.
4. Place the shell script provided here, "fastq_pipeline.sh", in the same directory. Edit this script as appropriate for the experiment and the computer directories.
   NOTE: this script contains extensive annotations explaining the steps. In brief, the script quality trims the reads, maps the reads to the genome, and generates a tab-delimited file containing the number of reads per gene for each sample.
5. Deposit the FASTQ files and raw read count data into NCBI's Gene Expression Omnibus before publication ²⁷. Follow the instructions for "Submitting high-throughput sequence data to GEO": https://www.ncbi.nlm.nih.gov/geo/info/seq.html.
6. Import the read count data into the R statistical environment and perform statistical analysis, using the R script provided here, "time_course_script.R". Edit this script as appropriate for the experiment and the computer directories.
   NOTE: This script contains extensive annotations explaining the steps. In brief, this script imports the read data, normalizes the reads, identifies genes that change significantly during the time course, performs PCA analysis, and graphs the expression of selected genes.

Results

The hypoxia time course and RNA-seq analysis were performed independently three times. To examine the reproducibility of the three replicates, gene expression data for all genes was analyzed using Principal Component Analysis (PCA). Figure 2 shows how the samples change over the first two principal components, which together represent 58.9% of the variability. This analysis indicated that each time course exhibits similar changes (as depicted by the similar s...

Discussion

In this study, the mRNA levels for all genes was measured during hypoxia in the yeast S. cerevisiae. The goal was to analyze how global gene expression changes due to growth in a controlled near-anoxic environment. Several steps were taken to ensure that the method described here was carefully controlled and reproducible. First, cells were exposed to a precisely defined hypoxic environment: 99.999% N₂ in rich media (YPD). Other studies of hypoxia have closed off the flask or tube to air

Materials

List of materials used in this article

Name	Company	Catalog Number	Comments
Enclosed dry incubator	Thermo Scientific	MaxQ 4000	Set at 30 °C. Modify the door to allow entry of one Tygon tube. Alternatively, use the New Brunswick G25 incubator, which contains a tube port. Do not use an open-air water shaker, as condensation will collect in the tubes between flasks, possibly cross-contaminating cultures.
Micro-analysis Filter Holder	Millipore	XX1002530	25mm diameter, stainless steel support, no. 5 perforated silicone stopper mounts in standard 125 mL filtering flask
Strong vacuum	Edwards	E-LAB 2	The “house” vacuum may be too weak. Alternatively, use an electric-power portable vacuum pump like the one listed here.
1,000 mL flask			To act as vacuum trap.
~2 foot lengths of Heavy Wall Vacuum Tubing, inner diameter 3/8 in, outer diameter 7/8 in	Tygon	38TT	Two pieces: the first connects vacuum to trap, and the second connects trap to filter system.
High-pressure N2 gas tank			99.999% purity, >1,000 psi, with a regulator and gas flow controller
Autoclaved 500 mL flask			Opening covered with aluminum foil. One for each yeast strain.
Autoclaved 250 mL flasks			Openings covered with aluminum foil. One for each time point plus two for water traps.
Flask stoppers (size 6, two holes with 5 mm diameter)			Sterilized with 70% ethanol. One for each flask.
Glass tubing, length 9 cm or 17 cm, inner diameter 2 mm, outer diameter 5 mm			Sterilized with 70% ethanol. Two tubes for each flask. Place into the holes of each stopper. See Figure 1 for placement of 9- vs 17- cm tubes.
~25 cm lengths of plastic tubing, inner diameter 5 mm	Tygon	E-3603	One piece for each flask. Sterilized with 70% ethanol.
Sterile filter discs	Millipore	HAWP02500	25 mm diameter, 0.45 µm pore size, one for each time point
Sterile dH2O (~100 mL)
1 mL cuvettes			For measuring OD600 (i.e., cell concentration)
50 mL sterile centrifuge tubes			One for each time point
Clean and sterile tweezers
liquid nitrogen			For freezing cells
acid-washed beads	Sigma	G8772	Keep at 4 °C for lysing cells
Qiagen RNeasy Mini Kit	Qiagen	74104	For RNA column purification
Qiagen RLT buffer			Prepare by adding 10 µL of β-Mercaptoethanol per 1 mL of RLT buffer, keep at 4 °C.
2 mL collection tubes	Qiagen		included in the Qiagen Rneasy Mini Kit
Buffer RPE	Qiagen		included in the Qiagen Rneasy Mini Kit
Buffer RW1	Qiagen		included in the Qiagen Rneasy Mini Kit
DNase I stock and working solutions	Qiagen	79254	The DNase I enzyme comes as lyophilized powder in a glass vial. Using a sterile needle and syringe, inject 550 µL of RNase-free water (provided in Qiagen kit) into the vial. Mix by gently inverting the bottle. To avoid denaturing the enzyme, do not vortex. Using a pipet, remove this stock solution from the vial and store in freezer (-20 °C) in single-use aliquots (80 µL each). The stock solution should not be thawed and refrozen.
Buffer RDD	Qiagen		included in the Qiagen DNase Kit
Ice cold 2 mL screw-cap tubes			For lysing cells during RNA extraction
bead mill homogenizer	Biospec Mini-Beadbeater-24	112011	Keep in cold room
Bacto Peptone	BD	DF0118	for liquid YPD media
Bacto Yeast Extract	BD	DF0886	for liquid YPD media
glucose	Fisher	D16	for liquid YPD media
Qubit assay tubes	Thermo Fisher	Q32856	for measuring nucleic acid concentration
Quant-iTTM dsDNA BR Assay Kit	Thermo Fisher	Q32853	for measuring nucleic acid concentration
Quant-iTTM RNA Assay Kit	Thermo Fisher	Q32855	for measuring nucleic acid concentration
Qubit Fluorometer	Thermo Fisher	Q33216	for measuring nucleic acid concentration
Commercial electrophoresis system	Agilent	Bioanalyzer 2100	for measuring nucleic acid quality
Next-generation sequencer	Illumina	HiSeq 2500	for sequencing libraries
automated liquid handling system	Wafergen	Apollo 324	for creating sequencing libraries
PrepX PolyA mRNA Isolation Kit	Wafergen	400047	for isolating mRNA from total RNA
PrepX RNA SEQ for Illumina Library Kit	Wafergen	400039	for creating strand-specific sequencing libraries from total RNA
Barcode Splitter			https://toolshed.g2.bx.psu.edu/repository?repository_id=7119c4f7a89efa57&changeset_revision=e7b7cdc1834d
Samtools, which includes the gzip command			http://www.htslib.org/download/
Trimmomatic			http://www.usadellab.org/cms/?page=trimmomatic
Bowtie2 (installed before TopHat)			http://bowtie-bio.sourceforge.net/bowtie2/index.shtml
TopHat			https://ccb.jhu.edu/software/tophat/index.shtml
HTSeq			http://www-huber.embl.de/HTSeq/doc/overview.html
R (installed before R Studio)			https://cran.rstudio.com
R Studio (free version)			https://www.rstudio.com/products/rstudio/download/